fix(langfuse): remove invalid user_id/session_id kwargs from start_as_current_observation

Langfuse V3 does not accept user_id/session_id on observation-level calls.
Moved to metadata dict in agent_runner, deep_agent, and agent_setup.

refactor(tests): fixture-based pattern for agent_runner_v2 eval tests

- cases.yaml + data/ fixtures under tests/fixtures/agent_runner_v2/
- pytest_generate_tests parametrizes test_eval_runner from YAML
- _resolve_projects() handles symbolic names and inline dicts
- _evaluate_case() centralizes all assertion logic
- --runner-dir CLI option for custom fixture folders

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

.env.example

@@ -39,6 +39,13 @@ QDRANT_URL=
 QDRANT_API_KEY=
 # For local Qdrant (homelab): QDRANT_URL=http://qdrant:6333
 
+# ── Langfuse (leave empty to disable observability) ───────────────────────────
+LANGFUSE_SECRET_KEY=
+LANGFUSE_PUBLIC_KEY=
+# LANGFUSE_HOST=https://cloud.langfuse.com        # EU (default)
+# LANGFUSE_HOST=https://us.cloud.langfuse.com     # US
+# LANGFUSE_HOST=http://localhost:3000             # Self-hosted
+
 # ── CORS ──────────────────────────────────────────────────────────────────────
 # Comma-separated list parsed by Settings (override default if needed)
 # CORS_ORIGINS=["app://.","http://localhost:3000"]

alembic/versions/a3b9c0d1e2f3_add_agent_config_to_local_agents.py

@@ -0,0 +1,31 @@
+"""add agent_config to local_agent_configs
+
+Revision ID: a3b9c0d1e2f3
+Revises: 9a1f2d0b6c7e
+Create Date: 2026-04-07 00:00:00.000000
+
+"""
+from __future__ import annotations
+
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision: str = "a3b9c0d1e2f3"
+down_revision: Union[str, None] = "9a1f2d0b6c7e"
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "local_agent_configs",
+        sa.Column("agent_config", sa.JSON(), nullable=True),
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("local_agent_configs", "agent_config")

app/api/routes/agent_setup.py

@@ -1,11 +1,11 @@
-"""Chatbot Journey — WS-based guided conversation to build an agent prompt_template.
+"""Chatbot Journey — WS-based guided conversation to build an AgentConfig.
 
 The journey is driven entirely through WebSocket frames (no REST endpoints).
 The device WS handler dispatches ``journey_start`` and ``journey_message``
 frames to the functions exported here.
 
 Journey flow:
-  1. FE sends ``journey_start`` frame with basic agent config (directory,
+  1. FE sends ``journey_start`` frame with basic agent info (directory,
      data_types, schedule).
   2. Server creates an in-memory session, sets up a WS executor so the
      setup LLM can use file-system tools, does a first directory scrape,
@@ -13,10 +13,11 @@ Journey flow:
   3. FE sends ``journey_message`` frames for each user reply.
   4. Server appends the user message, calls the LLM (which may read files
      via tools), and sends back a ``journey_reply``.
-  5. After 3-5 turns the LLM wraps up by emitting a ``prompt_template``
-     block delimited by ``PROMPT_TEMPLATE_START`` / ``PROMPT_TEMPLATE_END``.
-  6. Server parses the block, sends ``journey_reply`` with ``done=True``
-     and the template.  FE stores it locally.
+  5. After 3-5 turns the LLM wraps up by emitting an ``AgentConfig`` JSON
+     block delimited by ``AGENT_CONFIG_START`` / ``AGENT_CONFIG_END``.
+  6. Server parses and validates the JSON with Pydantic, sends
+     ``journey_reply`` with ``done=True`` and the serialised config.
+     FE stores it locally.
 """
 
 from __future__ import annotations
@@ -31,7 +32,10 @@ from typing import Any
 from langchain_core.messages import AIMessage, HumanMessage, SystemMessage, ToolMessage
 
 from app.agents.filesystem_agent import FILESYSTEM_TOOLS
+from app.config.settings import settings
+from app.core.langfuse_client import compile_prompt, extract_usage, get_langfuse, get_prompt_or_fallback
 from app.core.llm import get_llm
+from app.schemas import AgentConfig
 
 logger = logging.getLogger(__name__)
 
@@ -39,9 +43,9 @@ logger = logging.getLogger(__name__)
 
 _SESSION_TTL_SECONDS: int = 1800  # 30 minutes
 
-# Sentinel strings used to delimit the LLM-produced prompt_template.
-_TEMPLATE_START = "PROMPT_TEMPLATE_START"
-_TEMPLATE_END = "PROMPT_TEMPLATE_END"
+# Sentinel strings used to delimit the LLM-produced AgentConfig JSON.
+_CONFIG_START = "AGENT_CONFIG_START"
+_CONFIG_END = "AGENT_CONFIG_END"
 
 # Minimum turns before we consider nudging the LLM to wrap up.
 _MIN_TURNS_BEFORE_NUDGE: int = 3
@@ -62,6 +66,7 @@ class JourneySession:
     data_types: list[str]
     history: list[dict[str, Any]] = field(default_factory=list)
     system_prompt: str = ""
+    langfuse_prompt: Any = None
     created_at: float = field(default_factory=time.monotonic)
 
     def is_expired(self) -> bool:
@@ -83,61 +88,76 @@ def get_journey_session(session_id: str, user_id: str) -> JourneySession | None:
     return s
 
 
-# ── System prompt builder ─────────────────────────────────────────────────
+# ── System prompt ─────────────────────────────────────────────────────────
 
-_SYSTEM_PROMPT_TEMPLATE = """\
+_JOURNEY_SYSTEM_PROMPT = """\
 You are a friendly assistant helping a freelancer configure a data-extraction agent.
-Your job is to understand exactly what data the user wants to extract from their
-local directory and produce a detailed prompt_template that a separate AI will use
-as its instruction set.
-
-The extraction agent already has this base behaviour built in:
-  - Reads each file using file-system tools.
-  - Creates records (tasks, notes, timelines, projects) via CRUD tools.
-  - Sets isAiSuggested=1 on every new record.
-  - Only extracts data explicitly present in the files — it never invents information.
-The user's custom prompt is appended AFTER this base behaviour, so focus on
-what to look for and how to map it — not on the general extraction mechanics.
+Your job is to understand what files the user has in their directory and produce a
+structured AgentConfig JSON that the extraction agent will use as its instruction set.
 
 You have access to file-system tools to explore the user's directory:
-- list_directory: to see folder structure
-- read_file_content: to peek at file contents
-- get_file_metadata: to check file info
+- list_directory: see folder structure and file names
+- read_file_content: peek at a file's content
+- get_file_metadata: check file size, extension, dates
 
 The user's configured directory is: {directory}
 Target data types: {data_types}
 
-IMPORTANT — project assignment is handled automatically by the main agent runner
-before the custom prompt is ever used.  You MUST NOT ask the user about projects,
-projectId, or how to link records to projects.  Never include projectId logic or
-project creation instructions in the generated prompt_template.
+## Your process
 
-Start by exploring the directory to understand its structure.  Then ask concise,
-focused questions one at a time.  Cover these topics (not necessarily in this order):
-  1. The type and format of the source content (confirmed by your exploration).
-  2. How fields should be mapped (e.g. filename → task title).
-  3. Priority or status rules (e.g. "urgent" keyword → high priority).
-  4. Any special handling, date extraction, or exclusions.
+### Step 1 — Explore the directory
+Use list_directory and read_file_content to understand what types of files are present
+(HTML emails, plain-text documents, CSVs, etc.).
 
-Once you reach 90% confidence, output the final prompt_template between these exact
-markers on their own lines:
+### Step 2 — Identify content types
+For each distinct file type found, decide:
+- A short id (e.g. "email_html", "plain_text", "csv")
+- Which preprocessing handler to use: "email_html" for HTML emails, "generic" for everything else
+- A human-readable label and optional detection_hint
 
-{template_start}
-<the complete extraction prompt here>
-{template_end}
+### Step 3 — Ask focused questions (one at a time)
+Cover these topics based on what you discovered:
+1. How to map content to entity types (task / note / timeline entry)
+2. Field mapping rules (e.g. email Subject → task title, filename → note title)
+3. Priority or status rules (e.g. "urgent" in subject → high priority)
+4. Date extraction (e.g. "by Friday" → dueDate)
+5. Exclusion rules (e.g. skip newsletters, skip files with no project match)
 
-The prompt_template must be a self-contained instruction for an AI that reads files
-and must perform CRUD operations using tools to create records.  It should specify:
-  - What entity types to create (tasks, notes, timelines) — never projects.
-  - How to map file content to record fields (camelCase: title, status, priority,
-    dueDate, content, etc.) — never include projectId.
-  - That isAiSuggested must be set to 1 on every new record.
-  - Concrete examples of mappings based on what you discovered in the directory.
+### Step 4 — Produce the AgentConfig JSON
+Once you are ≥ 90% confident, output the final config between these exact markers
+(each on its own line):
+
+{config_start}
+{{
+  "content_types": [
+    {{
+      "id": "email_html",
+      "label": "Email HTML",
+      "detection_hint": "HTML file with From/To/Subject headers",
+      "preprocessing": "email_html",
+      "extraction_prompt": "Detailed extraction instructions for this content type..."
+    }}
+  ],
+  "global_rules": [
+    "If the file cannot be matched to any project, do not create any entity."
+  ],
+  "data_types": {data_types_json}
+}}
+{config_end}
+
+## Rules for the extraction_prompt field
+- Describe when to create a task vs note vs timeline entry (be specific and concrete)
+- Include field mapping rules based on what you found in the directory
+- Include priority/status/date rules if applicable
+- Do NOT include projectId logic — the runner handles project assignment automatically
+- Do NOT mention isAiSuggested — the runner always sets it to 1
+
+## Constraints
+- Never ask about projects, projectId, or how to link records to projects
+- Never include projectId or project creation logic in the generated config
+- Keep asking questions until ≥ 90% confident, then output the JSON immediately
 
 {existing_section}\
-Keep asking clarifying questions until you are at least 90% confident you have
-enough information to generate an accurate prompt_template.  Once you reach that
-confidence level, stop asking and produce the final template immediately.
 Begin by exploring the directory, then ask your first question.\
 """
 
@@ -145,33 +165,53 @@ Begin by exploring the directory, then ask your first question.\
 def _build_system_prompt(
     directory: str,
     data_types: list[str],
-    existing_template: str | None = None,
-) -> str:
+    existing_config: str | None = None,
+) -> tuple[str, Any]:
+    """Return ``(compiled_system_prompt, langfuse_prompt_obj_or_None)``."""
     existing_section = (
-        f"\nThe user already has the following prompt_template — refine it based on their answers:\n"
-        f"---\n{existing_template}\n---\n"
-        if existing_template
+        "\nThe user already has the following AgentConfig — refine it based on their answers:\n"
+        f"```json\n{existing_config}\n```\n"
+        if existing_config
         else ""
     )
-    return _SYSTEM_PROMPT_TEMPLATE.format(
+    template, prompt_obj = get_prompt_or_fallback(
+        "journey_system", _JOURNEY_SYSTEM_PROMPT
+    )
+    compiled = compile_prompt(
+        template,
+        prompt_obj,
         directory=directory,
         data_types=", ".join(data_types),
-        template_start=_TEMPLATE_START,
-        template_end=_TEMPLATE_END,
+        data_types_json=json.dumps(data_types),
+        config_start=_CONFIG_START,
+        config_end=_CONFIG_END,
         existing_section=existing_section,
     )
+    return compiled, prompt_obj
 
 
-# ── Template extraction ───────────────────────────────────────────────────
+# ── AgentConfig extraction ────────────────────────────────────────────────
 
 
-def _extract_template(text: str) -> str | None:
-    """Return the text between PROMPT_TEMPLATE_START and PROMPT_TEMPLATE_END, or None."""
-    if _TEMPLATE_START not in text or _TEMPLATE_END not in text:
+def _extract_agent_config(text: str) -> str | None:
+    """Return validated AgentConfig JSON string from between markers, or None.
+
+    Parses the JSON with Pydantic to ensure it conforms to the schema before
+    returning.  Returns None if markers are absent or JSON is invalid.
+    """
+    if _CONFIG_START not in text or _CONFIG_END not in text:
+        return None
+    start_idx = text.index(_CONFIG_START) + len(_CONFIG_START)
+    end_idx = text.index(_CONFIG_END)
+    raw = text[start_idx:end_idx].strip()
+    if not raw:
+        return None
+    try:
+        parsed = AgentConfig.model_validate_json(raw)
+        return parsed.model_dump_json()
+    except Exception as exc:
+        logger.warning("agent_setup: failed to parse AgentConfig JSON: %s", exc)
         return None
-    start_idx = text.index(_TEMPLATE_START) + len(_TEMPLATE_START)
-    end_idx = text.index(_TEMPLATE_END)
-    return text[start_idx:end_idx].strip() or None
 
 
 # ── LLM call with tool support ───────────────────────────────────────────
@@ -199,12 +239,17 @@ async def _call_llm_with_tools(
     system_prompt: str,
     history: list[dict[str, Any]],
     tools: list[Any],
+    *,
+    user_id: str = "",
+    session_id: str = "",
+    langfuse_prompt: Any = None,
 ) -> str:
     """Build LangChain messages from history and invoke the LLM with tools.
 
     Handles tool-calling loops: if the LLM calls tools, execute them and
     continue until a final text response is produced.
     """
+    lf = get_langfuse()
     messages: list[Any] = [SystemMessage(content=system_prompt)]
     for turn in history:
         if turn["role"] == "user":
@@ -216,38 +261,75 @@ async def _call_llm_with_tools(
     llm_with_tools = llm.bind_tools(tools)
     tool_map = {tool_def.name: tool_def for tool_def in tools}
 
-    for _ in range(_MAX_TOOL_STEPS):
-        response: AIMessage = await llm_with_tools.ainvoke(messages)
-        messages.append(response)
+    _span_ctx = (
+        lf.start_as_current_observation(
+            as_type="span",
+            name="journey-setup",
+            metadata={"user_id": user_id or None, "session_id": session_id or None},
+            input=history[-1]["content"] if history else "",
+        )
+        if lf else None
+    )
+    _span = _span_ctx.__enter__() if _span_ctx else None
 
-        if not response.tool_calls:
-            return _as_text(response.content)
-
-        for call in response.tool_calls:
-            call_name = str(call.get("name", ""))
-            call_args = call.get("args", {})
-            logger.info(
-                "agent_setup: journey tool_call name=%s args=%s",
-                call_name,
-                json.dumps(call_args, ensure_ascii=True)[:500],
+    try:
+        for _ in range(_MAX_TOOL_STEPS):
+            _gen_ctx = (
+                lf.start_as_current_observation(
+                    as_type="generation",
+                    name="journey-setup-llm",
+                    model=settings.LLM_MODEL,
+                    prompt=langfuse_prompt,
+                    input=messages,
+                )
+                if lf else None
             )
+            _gen = _gen_ctx.__enter__() if _gen_ctx else None
+            response: AIMessage = await llm_with_tools.ainvoke(messages)
+            if _gen_ctx:
+                _gen.update(output=_as_text(response.content), usage=extract_usage(response))
+                _gen_ctx.__exit__(None, None, None)
 
-            tool_fn = tool_map.get(call_name)
-            if tool_fn is None:
-                tool_output = f"Unknown tool: {call_name}"
-            else:
-                tool_output = await tool_fn.ainvoke(call_args)
+            messages.append(response)
 
-            logger.info(
-                "agent_setup: journey tool_result name=%s output=%s",
-                call_name,
-                str(tool_output)[:800],
-            )
-            messages.append(ToolMessage(content=str(tool_output), tool_call_id=call["id"]))
+            if not response.tool_calls:
+                if _span:
+                    _span.update(output=_as_text(response.content))
+                return _as_text(response.content)
 
-    # Fallback: exceeded max steps.
-    final = await llm.ainvoke(messages)
-    return _as_text(final.content)
+            for call in response.tool_calls:
+                call_name = str(call.get("name", ""))
+                call_args = call.get("args", {})
+                logger.info(
+                    "agent_setup: journey tool_call name=%s args=%s",
+                    call_name,
+                    json.dumps(call_args, ensure_ascii=True)[:500],
+                )
+
+                tool_fn = tool_map.get(call_name)
+                if tool_fn is None:
+                    tool_output = f"Unknown tool: {call_name}"
+                else:
+                    tool_output = await tool_fn.ainvoke(call_args)
+
+                logger.info(
+                    "agent_setup: journey tool_result name=%s output=%s",
+                    call_name,
+                    str(tool_output)[:800],
+                )
+                messages.append(ToolMessage(content=str(tool_output), tool_call_id=call["id"]))
+
+        # Fallback: exceeded max steps.
+        final = await llm.ainvoke(messages)
+        final_text = _as_text(final.content)
+        if _span:
+            _span.update(output=final_text)
+        return final_text
+    finally:
+        if _span_ctx:
+            _span_ctx.__exit__(None, None, None)
+        if lf:
+            lf.flush()
 
 
 # ── Journey handlers (called from device_ws.py) ──────────────────────────
@@ -265,12 +347,12 @@ async def handle_journey_start(
     agent_type = frame.get("agent_type", "local")
     directory = frame.get("directory", "")
     data_types = frame.get("data_types", [])
-    existing_template = frame.get("existing_template")
+    existing_config = frame.get("existing_config")
 
     # Use the session_id provided by the FE so the reply matches the
     # listener key; fall back to a generated one if absent.
     session_id = frame.get("session_id") or str(uuid.uuid4())
-    system_prompt = _build_system_prompt(directory, data_types, existing_template)
+    system_prompt, langfuse_prompt = _build_system_prompt(directory, data_types, existing_config)
 
     session = JourneySession(
         session_id=session_id,
@@ -279,12 +361,11 @@ async def handle_journey_start(
         directory=directory,
         data_types=data_types,
         system_prompt=system_prompt,
+        langfuse_prompt=langfuse_prompt,
     )
 
-    # The LLM will explore the directory using FILESYSTEM_TOOLS via the
-    # ws_context executor (already set by the WS handler before calling us).
-    # Seed with an initial user message — some providers (e.g. GitHub Copilot)
-    # require at least one user/input message to be present.
+    # Seed with an initial user message — some providers require at least one
+    # user/input message to be present.
     seed_history: list[dict[str, Any]] = [
         {"role": "user", "content": "Hi, I'm ready to set up my agent. Please explore my directory and ask me your first question."},
     ]
@@ -292,6 +373,9 @@ async def handle_journey_start(
         system_prompt=system_prompt,
         history=seed_history,
         tools=list(FILESYSTEM_TOOLS),
+        user_id=user_id,
+        session_id=session_id,
+        langfuse_prompt=langfuse_prompt,
     )
 
     session.history.extend(seed_history)
@@ -305,14 +389,14 @@ async def handle_journey_start(
         directory,
     )
 
-    # Check if the LLM produced the template on the first turn (unlikely but possible).
-    prompt_template = _extract_template(ai_reply)
-    done = prompt_template is not None
+    # Check if the LLM produced the config on the first turn (unlikely but possible).
+    agent_config = _extract_agent_config(ai_reply)
+    done = agent_config is not None
 
     display_message = ai_reply
     if done:
         display_message = (
-            ai_reply[: ai_reply.index(_TEMPLATE_START)].strip()
+            ai_reply[: ai_reply.index(_CONFIG_START)].strip()
             or "Here is your agent configuration. You can save it or continue refining."
         )
         _sessions.pop(session_id, None)
@@ -322,7 +406,7 @@ async def handle_journey_start(
         "session_id": session_id,
         "message": display_message,
         "done": done,
-        "prompt_template": prompt_template,
+        "agent_config": agent_config,
     }
 
 
@@ -345,7 +429,7 @@ async def handle_journey_message(
             "session_id": session_id,
             "message": "Journey session not found or expired. Please start a new setup.",
             "done": True,
-            "prompt_template": None,
+            "agent_config": None,
         }
 
     # Append user turn.
@@ -356,22 +440,24 @@ async def handle_journey_message(
         system_prompt=session.system_prompt,
         history=session.history,
         tools=list(FILESYSTEM_TOOLS),
+        user_id=session.user_id,
+        session_id=session_id,
+        langfuse_prompt=session.langfuse_prompt,
     )
 
     session.history.append({"role": "assistant", "content": ai_reply})
 
-    # Check if the LLM produced the final template.
-    prompt_template = _extract_template(ai_reply)
-    done = prompt_template is not None
+    # Check if the LLM produced the final config.
+    agent_config = _extract_agent_config(ai_reply)
+    done = agent_config is not None
 
-    # If the LLM didn't produce a template, nudge it once it has asked enough
-    # questions (>= _MIN_TURNS_BEFORE_NUDGE) or hits the hard safety cap.
+    # If the LLM didn't produce a config, nudge it once it hits the hard safety cap.
     if not done:
         turns = sum(1 for t in session.history if t["role"] == "user")
         if turns >= _MAX_TURNS:
             nudge_content = (
                 "[System: You have enough information. Please generate the final "
-                f"prompt_template now, wrapped in {_TEMPLATE_START} / {_TEMPLATE_END} markers.]"
+                f"AgentConfig JSON now, wrapped in {_CONFIG_START} / {_CONFIG_END} markers.]"
             )
             session.history.append({"role": "user", "content": nudge_content})
 
@@ -379,19 +465,22 @@ async def handle_journey_message(
                 system_prompt=session.system_prompt,
                 history=session.history,
                 tools=list(FILESYSTEM_TOOLS),
+                user_id=session.user_id,
+                session_id=session_id,
+                langfuse_prompt=session.langfuse_prompt,
             )
             session.history.append({"role": "assistant", "content": nudge_reply})
 
-            prompt_template = _extract_template(nudge_reply)
-            if prompt_template is not None:
+            agent_config = _extract_agent_config(nudge_reply)
+            if agent_config is not None:
                 done = True
                 ai_reply = nudge_reply
 
     display_message = ai_reply
     if done:
         display_message = (
-            ai_reply[: ai_reply.index(_TEMPLATE_START)].strip()
-            if _TEMPLATE_START in ai_reply
+            ai_reply[: ai_reply.index(_CONFIG_START)].strip()
+            if _CONFIG_START in ai_reply
             else "Here is your agent configuration. You can save it or continue refining."
         )
         _sessions.pop(session_id, None)
@@ -402,5 +491,5 @@ async def handle_journey_message(
         "session_id": session_id,
         "message": display_message,
         "done": done,
-        "prompt_template": prompt_template,
+        "agent_config": agent_config,
     }

app/config/settings.py

@@ -52,6 +52,10 @@ class Settings(BaseSettings):
 
     CORS_ORIGINS: list[str] = ["app://.", "http://localhost:3000", "http://localhost:5173"]
 
+    LANGFUSE_SECRET_KEY: str = ""
+    LANGFUSE_PUBLIC_KEY: str = ""
+    LANGFUSE_HOST: str = "https://cloud.langfuse.com"
+
     ENV: Literal["dev", "prod"] = "dev"
 
     model_config = SettingsConfigDict(env_file=".env", env_file_encoding="utf-8")

app/core/agent_runner.py

@@ -2,12 +2,12 @@
 
 Drives two agent types:
 
-* **Local directory agent** — two-step execution per file:
-  Step 1 (Classification) uses code to fetch all projects and asks the LLM
-  to identify which project the file belongs to and which domains are relevant.
-  Step 2 (Processing) fetches existing entities for that project/domains via
-  code and runs an LLM with tools — existing data in context enforces
-  update-first naturally.
+* **Local directory agent** — V2 unified flow per file:
+  Phase A (Detect + Preprocess, zero LLM): Python detects the content type
+  and strips markup/noise, producing clean text + metadata.
+  Phase B (Single LLM call with tools): the LLM identifies the project,
+  checks for duplicates via list_* tools, and creates/updates records.
+  ``items_created`` is counted from ``create_*`` tool calls.
 
 * **Cloud connector agent** — fetches data from third-party APIs (Gmail,
   Teams, Outlook) and pushes extracted items to Electron.
@@ -29,6 +29,7 @@ from __future__ import annotations
 import asyncio
 import json
 import logging
+import os
 import uuid
 from datetime import datetime, timedelta, timezone
 from typing import Any
@@ -42,8 +43,11 @@ from app.agents.note_agent import NOTE_TOOLS
 from app.agents.project_agent import PROJECT_TOOLS
 from app.agents.task_agent import TASK_TOOLS
 from app.agents.timeline_agent import TIMELINE_TOOLS
+from app.config.settings import settings
 from app.core.device_manager import DeviceConnectionManager
+from app.core.langfuse_client import compile_prompt, extract_usage, get_langfuse, get_prompt_or_fallback
 from app.core.llm import get_llm
+from app.core.preprocessors import detect_content_type, preprocess
 from app.core.ws_context import clear_client_executor, execute_on_client, set_client_executor
 from app.db import async_session
 from app.models import AgentRunLog, CloudAgentConfig, LocalAgentConfig
@@ -79,88 +83,43 @@ _DATA_TYPE_TOOLS: dict[str, list[Any]] = {
     "timelines": TIMELINE_TOOLS,
 }
 
-# ── Step 1: Classification prompt ─────────────────────────────────────────
+# ── V2: Unified processing prompt (hot-swappable via Langfuse "unified_processing") ──
 
-_DOMAIN_DESCRIPTIONS: dict[str, str] = {
-    "tasks": (
-        "Action items, to-dos, deliverables — anything that describes work to be done, "
-        "assigned to someone, or tracked with a due date or status."
-    ),
-    "notes": (
-        "Documentation, meeting notes, summaries, reference material — "
-        "written content meant to be read and referenced rather than acted on."
-    ),
-    "timelines": (
-        "Project milestones, deadlines, scheduled events — "
-        "specific dates that mark a point in the progress of a project."
-    ),
-    "projects": (
-        "High-level project entities — only relevant if the file clearly introduces "
-        "a new project or updates the scope of an existing one."
-    ),
-}
-
-_STEP1_SYSTEM_PROMPT = """\
-You are a file classifier for a freelance project management tool.
-
-Your job is to match a file to an existing project and identify which data domains to extract.
-
-## Project matching rules (STRICT — follow in order)
-
-1. Search the file content for any mention of a project name, client name, acronym, or topic
-   that overlaps with the existing projects listed below.
-2. The match does NOT need to be exact — partial name, abbreviation, or topic similarity is enough.
-3. STRONGLY PREFER matching an existing project. Only return "new" as an absolute last resort
-   when the file has zero meaningful connection to any listed project.
-4. When in doubt, pick the closest match from the list.
-
-## Response format
-
-Respond ONLY with a JSON object — no markdown, no explanation:
-
-{{"project_id": "<exact id from the list below, or new>", "new_project_name": "<concise 2-5 word name, only when project_id is new>", "domains": ["tasks", "notes"]}}
-
-## Domain definitions (only consider domains in the allowed list)
-
-{domain_definitions}
-
-## Existing projects
-
-{projects_list}
-"""
-
-# ── Step 2: Processing prompt ─────────────────────────────────────────────
-
-_PROCESSING_SYSTEM_PROMPT = """\
+_UNIFIED_PROCESSING_PROMPT = """\
 You are a data extraction assistant for a freelance project management tool.
 
-Your task: extract structured data from the file content and persist it using the available tools.
+## Your process (follow this exact order)
 
-## Mandatory process — follow this order for EVERY item you extract
+### 1. Identify the project
+File: {filename}
+{metadata_section}
 
-1. READ the existing records listed below for the relevant domain.
-2. SEARCH for a match by title, topic, or semantic similarity.
-3. If a match exists → call the update_* tool with the existing record's id.
-4. If no match exists → call the create_* tool and set isAiSuggested=1.
+Existing projects:
+{projects_list}
 
-NEVER call create_* without first checking the existing records.
-NEVER duplicate a record that already exists under a different wording.
+Match this file to an existing project using the filename and content clues.
+If no project matches, {no_match_behavior}.
 
-## Existing records (source of truth)
+### 2. Check existing records
+Once you identify the project, use list_tasks / list_notes / list_timelines
+(filtered by projectId) to see what already exists.
+NEVER create a record that already exists under the same or similar title.
 
-{existing_context}
+### 3. Extract and create / update
+{extraction_rules}
 
-## Context
-
-Project: {project_context}
-Domains to extract: {data_types}
-
-{custom_prompt_section}
+### Rules
+- Set isAiSuggested=1 on every new record.
+- Set projectId on every record (use the id from the project list above).
+- Update existing records when a match is found by title or topic.
+- Do NOT invent data — only extract what is clearly stated in the content.
+- Target entity types: {data_types}.
+{global_rules}
 """
 
 # ── Cloud processing prompt (kept separate for cloud agent) ───────────────
 
-_CLOUD_PROCESSING_PROMPT = """\
+_BATCH_CLOUD_PROCESSING_PROMPT = """\
 You are a data extraction and management assistant for a freelance project
 management tool.
 
@@ -268,8 +227,17 @@ async def _run_agent_with_tools(
     user_message: str,
     tools: list[Any],
     max_steps: int,
+    user_id: str = "",
+    langfuse_prompt: Any = None,
+    agent_name: str = "batch-agent",
+    _tool_calls_out: list[str] | None = None,
 ) -> str:
-    """Run an LLM agent with tool-calling, returning the final text response."""
+    """Run an LLM agent with tool-calling, returning the final text response.
+
+    If *_tool_calls_out* is provided, the name of every tool called during the
+    run is appended to it (used by the caller to count ``create_*`` calls).
+    """
+    lf = get_langfuse()
     llm = get_llm()
     llm_with_tools = llm.bind_tools(tools)
     messages: list[Any] = [
@@ -279,38 +247,79 @@ async def _run_agent_with_tools(
 
     tool_map = {tool_def.name: tool_def for tool_def in tools}
 
-    for _ in range(max_steps):
-        response: AIMessage = await llm_with_tools.ainvoke(messages)
-        messages.append(response)
+    _span_ctx = (
+        lf.start_as_current_observation(
+            as_type="span",
+            name=agent_name,
+            metadata={"user_id": user_id} if user_id else None,
+            input=user_message,
+        )
+        if lf else None
+    )
+    _span = _span_ctx.__enter__() if _span_ctx else None
 
-        if not response.tool_calls:
-            return _as_text(response.content)
-
-        for call in response.tool_calls:
-            call_id = str(call.get("id", ""))
-            call_name = str(call.get("name", ""))
-            call_args = call.get("args", {})
-            logger.info(
-                "agent_runner: tool_call name=%s args=%s",
-                call_name,
-                json.dumps(call_args, ensure_ascii=True)[:800],
+    try:
+        for _ in range(max_steps):
+            _gen_ctx = (
+                lf.start_as_current_observation(
+                    as_type="generation",
+                    name=f"{agent_name}-llm",
+                    model=settings.LLM_MODEL,
+                    prompt=langfuse_prompt,
+                    input=messages,
+                )
+                if lf else None
             )
+            _gen = _gen_ctx.__enter__() if _gen_ctx else None
+            response: AIMessage = await llm_with_tools.ainvoke(messages)
+            if _gen_ctx:
+                _gen.update(output=_as_text(response.content), usage=extract_usage(response))
+                _gen_ctx.__exit__(None, None, None)
 
-            tool_fn = tool_map.get(call_name)
-            if tool_fn is None:
-                tool_output = f"Unknown tool: {call_name}"
-            else:
-                tool_output = await tool_fn.ainvoke(call_args)
+            messages.append(response)
 
-            logger.info(
-                "agent_runner: tool_result name=%s output=%s",
-                call_name,
-                str(tool_output)[:200],
-            )
-            messages.append(ToolMessage(content=str(tool_output), tool_call_id=call["id"]))
+            if not response.tool_calls:
+                final_text = _as_text(response.content)
+                if _span:
+                    _span.update(output=final_text)
+                return final_text
 
-    final = await llm.ainvoke(messages)
-    return _as_text(final.content)
+            for call in response.tool_calls:
+                call_id = str(call.get("id", ""))
+                call_name = str(call.get("name", ""))
+                call_args = call.get("args", {})
+                logger.info(
+                    "agent_runner: tool_call name=%s args=%s",
+                    call_name,
+                    json.dumps(call_args, ensure_ascii=True)[:800],
+                )
+
+                if _tool_calls_out is not None:
+                    _tool_calls_out.append(call_name)
+
+                tool_fn = tool_map.get(call_name)
+                if tool_fn is None:
+                    tool_output = f"Unknown tool: {call_name}"
+                else:
+                    tool_output = await tool_fn.ainvoke(call_args)
+
+                logger.info(
+                    "agent_runner: tool_result name=%s output=%s",
+                    call_name,
+                    str(tool_output)[:200],
+                )
+                messages.append(ToolMessage(content=str(tool_output), tool_call_id=call["id"]))
+
+        final = await llm.ainvoke(messages)
+        final_text = _as_text(final.content)
+        if _span:
+            _span.update(output=final_text)
+        return final_text
+    finally:
+        if _span_ctx:
+            _span_ctx.__exit__(None, None, None)
+        if lf:
+            lf.flush()
 
 
 # ── Tool list builder ─────────────────────────────────────────────────────
@@ -479,83 +488,66 @@ def _format_entities_for_context(domain: str, rows: list[dict]) -> str:
     return f"Existing {domain}:\n" + "\n".join(lines)
 
 
-# ── Step 1: LLM file classifier ───────────────────────────────────────────
+# ── V2 helper functions ───────────────────────────────────────────────────
 
 
-async def _classify_file(
-    file_path: str,
-    file_content: str,
-    projects: list[dict],
-    config_data_types: list[str],
-) -> tuple[str, list[str], str | None]:
-    """Call the LLM to classify a file by project and relevant domains.
-
-    Returns ``(project_id_or_"new", domains, new_project_name_or_None)``.
-    - ``project_id`` is an existing project UUID, or ``"new"`` when no match found.
-    - ``new_project_name`` is only set when ``project_id == "new"``.
-    Falls back to ``("new", config_data_types, None)`` on any error.
-    """
-    fallback: tuple[str, list[str], str | None] = ("new", list(config_data_types), None)
-
-    if not file_content.strip():
-        return fallback
-
-    valid_project_ids = {p["id"] for p in projects}
-
-    def _fmt_project(p: dict) -> str:
+def _format_projects(projects: list[dict]) -> str:
+    """Format the project list for the unified system prompt."""
+    if not projects:
+        return "  (no projects yet)"
+    lines: list[str] = []
+    for p in projects:
         summary = (p.get("aiSummary") or p.get("ai_summary") or "").strip()
         summary_part = f" — {summary[:100]}" if summary else ""
-        return f"  - id={p['id']} | name={p.get('name', '')} | status={p.get('status', '')}{summary_part}"
+        lines.append(
+            f"  - id={p['id']} | name={p.get('name', '')} | "
+            f"status={p.get('status', '')}{summary_part}"
+        )
+    return "\n".join(lines)
 
-    projects_list = "\n".join(_fmt_project(p) for p in projects) or "  (none yet)"
 
-    domain_definitions = "\n".join(
-        f"  - {d}: {_DOMAIN_DESCRIPTIONS[d]}"
-        for d in config_data_types
-        if d in _DOMAIN_DESCRIPTIONS
+def _format_metadata(metadata: dict) -> str:
+    """Format preprocessor metadata as a compact context block."""
+    if not metadata:
+        return ""
+    parts: list[str] = []
+    for key in ("subject", "from", "to", "date"):
+        if metadata.get(key):
+            parts.append(f"{key.capitalize()}: {metadata[key]}")
+    # any remaining keys
+    for key, val in metadata.items():
+        if key not in ("subject", "from", "to", "date") and val:
+            parts.append(f"{key}: {val}")
+    return "\n".join(parts)
+
+
+def _get_extraction_rules(agent_config: dict, content_type: str) -> str:
+    """Return the extraction_prompt for *content_type* from *agent_config*.
+
+    Falls back to a generic instruction when the type is not configured.
+    """
+    for ct in agent_config.get("content_types", []):
+        if ct.get("id") == content_type:
+            prompt = ct.get("extraction_prompt", "").strip()
+            if prompt:
+                return prompt
+    return (
+        "Extract relevant information as tasks (action items), notes "
+        "(informational content), or timelines (dated events)."
     )
 
-    system = _STEP1_SYSTEM_PROMPT.format(
-        domain_definitions=domain_definitions,
-        projects_list=projects_list,
-    )
 
-    llm = get_llm()
-    try:
-        response = await llm.ainvoke([
-            SystemMessage(content=system),
-            HumanMessage(content=f"File: {file_path}\n\nContent:\n{file_content[:4000]}"),
-        ])
-        raw = _as_text(response.content).strip()
-        # Strip markdown fences if the model wraps the JSON.
-        if raw.startswith("```"):
-            raw = raw.split("```")[1]
-            if raw.startswith("json"):
-                raw = raw[4:]
-        parsed = json.loads(raw.strip())
-        raw_project_id: str = str(parsed.get("project_id") or "new")
-        # Reject hallucinated UUIDs — only accept ids that exist in the fetched list.
-        project_id = raw_project_id if raw_project_id in valid_project_ids else "new"
-        new_project_name: str | None = (
-            str(parsed["new_project_name"]).strip() or None
-            if project_id == "new" and parsed.get("new_project_name")
-            else None
-        )
-        domains: list[str] = [
-            d for d in parsed.get("domains", [])
-            if d in config_data_types
-        ]
-        if not domains:
-            domains = list(config_data_types)
-        return project_id, domains, new_project_name
-    except Exception as exc:
-        logger.warning(
-            "agent_runner: step1 classification failed for %r: %s", file_path, exc
-        )
-        return fallback
+def _get_no_match_behavior(agent_config: dict) -> str:
+    """Derive the 'no project match' instruction from global_rules."""
+    rules = agent_config.get("global_rules", [])
+    for rule in rules:
+        lower = rule.lower()
+        if "no project" in lower or "no match" in lower or "skip" in lower:
+            return rule
+    return "create a new project with a concise name derived from the file content"
 
 
-# ── Local agent runner (two-step per file) ────────────────────────────────
+# ── Local agent runner (V2 — unified per-file flow) ───────────────────────
 
 
 async def run_local_agent(
@@ -565,16 +557,17 @@ async def run_local_agent(
     device_mgr: DeviceConnectionManager,
     run_context: dict | None = None,
 ) -> None:
-    """Execute a local directory agent run using a two-step approach per file.
+    """Execute a local directory agent run — V2 unified flow.
 
-    Step 1 — Classification (code + 1 LLM call per file, no tools):
-        Code scans directories and fetches all projects via WS.
-        For each file, LLM identifies the project and relevant domains.
+    Phase A — Detect + Preprocess (zero LLM, per file):
+        Python detects the content type from filename + content patterns and
+        runs the appropriate handler (e.g. email_html) to produce clean text
+        and structured metadata.
 
-    Step 2 — Processing (code + 1 LLM call per file, with tools):
-        Code fetches existing entities for the identified project/domains.
-        LLM receives file content + existing entities in context and uses
-        tools to update existing records or create new ones.
+    Phase B — Single LLM call with tools (per file):
+        One LLM call handles project identification, duplicate checking, and
+        record creation/update.  ``create_*`` tool calls are counted to
+        produce the accurate ``items_created`` metric.
     """
     run_id = run_log.id
     agent_id = (run_context or {}).get("agent_id") or config.id
@@ -609,12 +602,8 @@ async def run_local_agent(
     errors: list[str] = []
     items_processed = 0
     items_created = 0
-
-    custom_section = (
-        f"User instructions:\n{config.prompt_template}"
-        if config.prompt_template
-        else ""
-    )
+    agent_config: dict = config.agent_config or {}
+    processing_tools = _build_processing_tools(config.data_types)
 
     try:
         # ── Code: scan directories ───────────────────────────────────
@@ -634,108 +623,82 @@ async def run_local_agent(
 
         # ── Code: fetch all projects once ────────────────────────────
         projects = await _fetch_projects()
+        projects_block = _format_projects(projects)
+
+        # Prompt template + Langfuse version linking (hot-swappable from UI).
+        unified_template, prompt_obj = get_prompt_or_fallback(
+            "unified_processing", _UNIFIED_PROCESSING_PROMPT
+        )
 
         for file_path in file_paths:
             try:
-                # Read file content via code.
+                # ── Phase A: read + detect + preprocess ─────────────
                 file_result = await execute_on_client(
                     action="read_file_content", data={"path": file_path}
                 )
-                file_content: str = file_result.get("content", "")
-                if not file_content:
-                    logger.debug("agent_runner: run=%s skipping empty file %r", run_id, file_path)
+                raw_content: str = file_result.get("content", "")
+                if not raw_content.strip():
+                    logger.debug(
+                        "agent_runner: run=%s skipping empty file %r", run_id, file_path
+                    )
                     continue
 
                 items_processed += 1
+                filename = os.path.basename(file_path)
+                content_type = detect_content_type(filename, raw_content)
+                preprocessed = preprocess(content_type, raw_content)
 
-                # Step 1 — classify file.
-                project_id, domains, new_project_name = await _classify_file(
-                    file_path=file_path,
-                    file_content=file_content,
-                    projects=projects,
-                    config_data_types=config.data_types,
-                )
                 logger.info(
-                    "agent_runner: run=%s file=%r → project=%s new_name=%r domains=%s",
-                    run_id,
-                    file_path,
-                    project_id,
-                    new_project_name,
-                    domains,
+                    "agent_runner: run=%s file=%r content_type=%s clean_len=%d",
+                    run_id, file_path, content_type, len(preprocessed.clean_text),
                 )
 
-                # Step 2 — resolve project_id via CODE, then fetch entities.
-                # Project creation is NEVER delegated to the Step 2 LLM.
-                if project_id == "new":
-                    proj_name = new_project_name or "Untitled Project"
-                    try:
-                        proj_result = await execute_on_client(
-                            action="insert",
-                            table="projects",
-                            data={"name": proj_name, "clientId": None},
-                        )
-                        created = proj_result.get("row", {})
-                        effective_project_id = created.get("id", "standalone")
-                        # Add to local list so subsequent files can match it.
-                        if "id" in created:
-                            projects.append(created)
-                        logger.info(
-                            "agent_runner: run=%s created project %r id=%s",
-                            run_id, proj_name, effective_project_id,
-                        )
-                    except Exception as exc:
-                        logger.warning(
-                            "agent_runner: run=%s failed to create project %r: %s",
-                            run_id, proj_name, exc,
-                        )
-                        effective_project_id = "standalone"
-                        proj_name = "unknown"
-                    project_context = (
-                        f"Project: {proj_name} (id: {effective_project_id}). "
-                        "Always set projectId to this id on every record you create."
-                    )
-                else:
-                    effective_project_id = project_id
-                    proj = next((p for p in projects if p["id"] == project_id), None)
-                    proj_name = proj.get("name", project_id) if proj else project_id
-                    project_context = (
-                        f"Project: {proj_name} (id: {project_id}). "
-                        "Always set projectId to this id on every record you create."
-                    )
+                # ── Phase B: single LLM call ─────────────────────────
+                extraction_rules = _get_extraction_rules(agent_config, content_type)
+                no_match_behavior = _get_no_match_behavior(agent_config)
+                global_rules_lines = "\n".join(
+                    f"- {r}" for r in agent_config.get("global_rules", [])
+                )
+                metadata_section = _format_metadata(preprocessed.metadata)
 
-                # "projects" domain is never passed to Step 2 — handled above in code.
-                domains = [d for d in domains if d != "projects"]
-
-                existing_blocks: list[str] = []
-                for domain in domains:
-                    rows = await _fetch_domain_entities(domain, effective_project_id)
-                    existing_blocks.append(_format_entities_for_context(domain, rows))
-
-                existing_context = "\n\n".join(existing_blocks)
-
-                system_prompt = _PROCESSING_SYSTEM_PROMPT.format(
-                    existing_context=existing_context,
-                    project_context=project_context,
-                    data_types=", ".join(domains),
-                    custom_prompt_section=custom_section,
+                system_prompt = compile_prompt(
+                    unified_template,
+                    prompt_obj,
+                    filename=filename,
+                    metadata_section=metadata_section,
+                    projects_list=projects_block,
+                    no_match_behavior=no_match_behavior,
+                    extraction_rules=extraction_rules,
+                    global_rules=global_rules_lines,
+                    data_types=", ".join(config.data_types),
                 )
 
-                processing_tools = _build_processing_tools(domains)
+                user_message = (
+                    f"Process this file and extract relevant information.\n\n"
+                    f"File: {file_path}\n\n"
+                    f"Content:\n{preprocessed.clean_text}"
+                )
 
+                file_tool_calls: list[str] = []
                 result_text = await _run_agent_with_tools(
                     system_prompt=system_prompt,
-                    user_message=(
-                        f"Process this file and extract relevant information.\n\n"
-                        f"File: {file_path}\n\nContent:\n{file_content}"
-                    ),
+                    user_message=user_message,
                     tools=processing_tools,
                     max_steps=_MAX_PROCESSING_STEPS,
+                    user_id=user_id,
+                    langfuse_prompt=prompt_obj,
+                    agent_name="unified-processor",
+                    _tool_calls_out=file_tool_calls,
                 )
+
+                file_created = sum(
+                    1 for name in file_tool_calls if name.startswith("create_")
+                )
+                items_created += file_created
+
                 logger.info(
-                    "agent_runner: run=%s file=%r result=%s",
-                    run_id,
-                    file_path,
-                    result_text[:200],
+                    "agent_runner: run=%s file=%r created=%d result=%s",
+                    run_id, file_path, file_created, result_text[:200],
                 )
 
             except Exception as exc:
@@ -767,10 +730,11 @@ async def run_local_agent(
         errors=errors,
     )
     logger.info(
-        "agent_runner: run=%s done status=%s processed=%d errors=%d",
+        "agent_runner: run=%s done status=%s processed=%d created=%d errors=%d",
         run_id,
         final_status,
         items_processed,
+        items_created,
         len(errors),
     )
 
@@ -928,7 +892,12 @@ async def run_cloud_agent(
                 continue
             items_processed += 1
 
-            processing_prompt = _CLOUD_PROCESSING_PROMPT.format(
+            cloud_template, cloud_prompt_obj = get_prompt_or_fallback(
+                "batch_cloud_processing", _BATCH_CLOUD_PROCESSING_PROMPT
+            )
+            processing_prompt = compile_prompt(
+                cloud_template,
+                cloud_prompt_obj,
                 data_types=", ".join(config.data_types),
                 project_context="Determine the appropriate project from the message context.",
                 file_list=f"Message from {config.provider} (id: {msg.id})",
@@ -941,6 +910,9 @@ async def run_cloud_agent(
                     user_message=f"Process this message content:\n\n{content_text[:8000]}",
                     tools=processing_tools,
                     max_steps=_MAX_PROCESSING_STEPS,
+                    user_id=user_id,
+                    langfuse_prompt=cloud_prompt_obj,
+                    agent_name="cloud-processor",
                 )
             except Exception as exc:
                 errors.append(f"LLM processing error for message {msg.id!r}: {exc}")

app/core/deep_agent.py

@@ -16,7 +16,9 @@ from app.agents.note_agent import NOTE_TOOLS
 from app.agents.project_agent import PROJECT_TOOLS
 from app.agents.task_agent import TASK_TOOLS
 from app.agents.timeline_agent import TIMELINE_TOOLS
+from app.core.langfuse_client import extract_usage, get_langfuse, get_prompt_or_fallback
 from app.core.llm import get_llm
+from app.config.settings import settings
 from app.core.memory_middleware import MemoryMiddleware
 from app.core.ws_context import clear_tool_result_collector, execute_on_client, set_tool_result_collector
 from app.db import async_session
@@ -26,7 +28,7 @@ logger = logging.getLogger(__name__)
 FloatingDomainType = Literal["task", "timeline", "project", "node"]
 FloatingDomainSection = Literal["task", "timeline", "note"]
 
-_HOME_SINGLE_AGENT_SYSTEM = (
+_HOME_SYSTEM_PROMPT = (
     "You are the home assistant with direct access to all tools: tasks, projects, notes, timelines, and memory tools. "
     "Always use tools for factual data retrieval before answering. "
     "When the user asks to remember, forget, or update what you know about them, use memory tools. "
@@ -39,7 +41,7 @@ _HOME_SINGLE_AGENT_SYSTEM = (
     "For upcoming tasks, after tag lines add a short recommendation based on due date and priority."
 )
 
-_FLOATING_SINGLE_AGENT_SYSTEM = (
+_FLOATING_SYSTEM_PROMPT = (
     "You are the floating assistant with direct access to all tools: tasks, projects, notes, timelines, and memory tools. "
     "Stay focused on the floating scope in context.scope and answer concisely. "
     "Return plain text only. Do not output XML/HTML-like tags such as <task>, <project>, <note>, <timeline>, or any bracketed id tag wrappers. "
@@ -48,7 +50,7 @@ _FLOATING_SINGLE_AGENT_SYSTEM = (
     "If context.context.resolved_project_id exists, use it as project_id for scoped list calls. "
 )
 
-_FLOATING_DOMAIN_CLASSIFIER_SYSTEM = (
+_FLOATING_DOMAIN_CLASSIFIER_PROMPT = (
     "You are a strict domain classifier for websocket floating requests. "
     "Return ONLY a JSON object with keys: type, id, section. "
     "Allowed type values: task, timeline, project, node. "
@@ -536,17 +538,31 @@ async def _infer_floating_domain(message: str, context: dict[str, Any]) -> dict[
 
     try:
         llm = get_llm()
-        response = await llm.ainvoke(
-            [
-                SystemMessage(content=_FLOATING_DOMAIN_CLASSIFIER_SYSTEM),
-                HumanMessage(
-                    content=(
-                        f"Message:\n{message}\n\n"
-                        f"Context:\n{json.dumps(classifier_context, ensure_ascii=True)}"
-                    )
-                ),
-            ]
+        classifier_messages = [
+            SystemMessage(content=_FLOATING_DOMAIN_CLASSIFIER_PROMPT),
+            HumanMessage(
+                content=(
+                    f"Message:\n{message}\n\n"
+                    f"Context:\n{json.dumps(classifier_context, ensure_ascii=True)}"
+                )
+            ),
+        ]
+        lf = get_langfuse()
+        _, classifier_prompt_obj = get_prompt_or_fallback(
+            "floating_domain_classifier", _FLOATING_DOMAIN_CLASSIFIER_PROMPT
         )
+        if lf:
+            with lf.start_as_current_observation(
+                as_type="generation",
+                name="floating-classifier",
+                model=settings.LLM_MODEL,
+                prompt=classifier_prompt_obj,
+                input=classifier_messages,
+            ) as gen:
+                response = await llm.ainvoke(classifier_messages)
+                gen.update(output=_as_text(response.content), usage=extract_usage(response))
+        else:
+            response = await llm.ainvoke(classifier_messages)
         parsed = _parse_json_object(_as_text(response.content))
         if parsed is not None:
             domain = _normalize_domain_payload(parsed, project_id)
@@ -571,8 +587,11 @@ async def _run_single_agent(
     message: str,
     context: dict[str, Any],
     max_steps: int = 6,
+    langfuse_prompt: Any = None,
+    agent_name: str = "agent",
 ) -> str:
     trace_id = _trace_id_from_context(context)
+    lf = get_langfuse()
     llm = get_llm()
     tools = _all_tools_for_user(user_id, trace_id)
     model_context = _context_for_model(context)
@@ -591,9 +610,36 @@ async def _run_single_agent(
     tool_calls_count = 0
     collected: list[dict[str, Any]] = []
     set_tool_result_collector(collected)
+
+    _span_ctx = (
+        lf.start_as_current_observation(
+            as_type="span",
+            name=agent_name,
+            metadata={"user_id": user_id, "session_id": trace_id},
+            input=message,
+        )
+        if lf else None
+    )
+    _span = _span_ctx.__enter__() if _span_ctx else None
+
     try:
         for _ in range(max_steps):
+            _gen_ctx = (
+                lf.start_as_current_observation(
+                    as_type="generation",
+                    name=f"{agent_name}-llm",
+                    model=settings.LLM_MODEL,
+                    prompt=langfuse_prompt,
+                    input=messages,
+                )
+                if lf else None
+            )
+            _gen = _gen_ctx.__enter__() if _gen_ctx else None
             response: AIMessage = await llm_with_tools.ainvoke(messages)
+            if _gen_ctx:
+                _gen.update(output=_as_text(response.content), usage=extract_usage(response))
+                _gen_ctx.__exit__(None, None, None)
+
             messages.append(response)
 
             if not response.tool_calls:
@@ -605,6 +651,8 @@ async def _run_single_agent(
                     tool_calls_count,
                     len(final_text),
                 )
+                if _span:
+                    _span.update(output=final_text)
                 return final_text
 
             tool_map = {tool_def.name: tool_def for tool_def in tools}
@@ -644,9 +692,15 @@ async def _run_single_agent(
             tool_calls_count,
             len(final_text),
         )
+        if _span:
+            _span.update(output=final_text)
         return final_text
     finally:
         clear_tool_result_collector()
+        if _span_ctx:
+            _span_ctx.__exit__(None, None, None)
+        if lf:
+            lf.flush()
 
 
 async def _run_single_agent_stream(
@@ -656,8 +710,11 @@ async def _run_single_agent_stream(
     message: str,
     context: dict[str, Any],
     max_steps: int = 6,
+    langfuse_prompt: Any = None,
+    agent_name: str = "agent",
 ) -> AsyncGenerator[tuple[str, Any], None]:
     trace_id = _trace_id_from_context(context)
+    lf = get_langfuse()
     llm = get_llm()
     tools = _all_tools_for_user(user_id, trace_id)
     model_context = _context_for_model(context)
@@ -677,9 +734,37 @@ async def _run_single_agent_stream(
     streamed_chars = 0
     collected: list[dict[str, Any]] = []
     set_tool_result_collector(collected)
+
+    _span_ctx = (
+        lf.start_as_current_observation(
+            as_type="span",
+            name=f"{agent_name}-stream",
+            metadata={"user_id": user_id, "session_id": trace_id},
+            input=message,
+        )
+        if lf else None
+    )
+    _span = _span_ctx.__enter__() if _span_ctx else None
+    streamed_text: list[str] = []
+
     try:
         for _ in range(max_steps):
+            _gen_ctx = (
+                lf.start_as_current_observation(
+                    as_type="generation",
+                    name=f"{agent_name}-llm",
+                    model=settings.LLM_MODEL,
+                    prompt=langfuse_prompt,
+                    input=messages,
+                )
+                if lf else None
+            )
+            _gen = _gen_ctx.__enter__() if _gen_ctx else None
             response: AIMessage = await llm_with_tools.ainvoke(messages)
+            if _gen_ctx:
+                _gen.update(output=_as_text(response.content), usage=extract_usage(response))
+                _gen_ctx.__exit__(None, None, None)
+
             messages.append(response)
 
             if not response.tool_calls:
@@ -688,6 +773,7 @@ async def _run_single_agent_stream(
                     token = _as_text(getattr(chunk, "content", ""))
                     if token:
                         streamed_chars += len(token)
+                        streamed_text.append(token)
                         emitted_any = True
                         yield "token", token
 
@@ -696,6 +782,7 @@ async def _run_single_agent_stream(
                     fallback_text = _as_text(response.content)
                     if fallback_text:
                         streamed_chars += len(fallback_text)
+                        streamed_text.append(fallback_text)
                         yield "token", fallback_text
                 logger.info(
                     "deep_agent: run_single_agent_stream_end trace=%s user=%s tool_calls=%d response_chars=%d",
@@ -704,6 +791,8 @@ async def _run_single_agent_stream(
                     tool_calls_count,
                     streamed_chars,
                 )
+                if _span:
+                    _span.update(output="".join(streamed_text))
                 return
 
             tool_map = {tool_def.name: tool_def for tool_def in tools}
@@ -738,6 +827,7 @@ async def _run_single_agent_stream(
             token = _as_text(getattr(chunk, "content", ""))
             if token:
                 streamed_chars += len(token)
+                streamed_text.append(token)
                 yield "token", token
         logger.info(
             "deep_agent: run_single_agent_stream_end trace=%s user=%s tool_calls=%d response_chars=%d fallback=1",
@@ -746,17 +836,28 @@ async def _run_single_agent_stream(
             tool_calls_count,
             streamed_chars,
         )
+        if _span:
+            _span.update(output="".join(streamed_text))
     finally:
         clear_tool_result_collector()
+        if _span_ctx:
+            _span_ctx.__exit__(None, None, None)
+        if lf:
+            lf.flush()
 
 
 async def run_home(user_id: str, message: str, context: dict[str, Any]) -> str:
     prepared_context = await _prepare_context(message, context)
+    system_prompt, langfuse_prompt = get_prompt_or_fallback(
+        "home_system", _HOME_SYSTEM_PROMPT
+    )
     response = await _run_single_agent(
         user_id=user_id,
-        system_prompt=_HOME_SINGLE_AGENT_SYSTEM,
+        system_prompt=system_prompt,
         message=message,
         context=prepared_context,
+        langfuse_prompt=langfuse_prompt,
+        agent_name="home-agent",
     )
     return _normalize_tagged_list_lines(response, message)
 
@@ -764,11 +865,16 @@ async def run_home(user_id: str, message: str, context: dict[str, Any]) -> str:
 async def run_floating(user_id: str, message: str, context: dict[str, Any]) -> tuple[str, dict[str, str | None]]:
     prepared_context = await _prepare_context(message, context)
     domain = await _infer_floating_domain(message, prepared_context)
+    system_prompt, langfuse_prompt = get_prompt_or_fallback(
+        "floating_system", _FLOATING_SYSTEM_PROMPT
+    )
     response = await _run_single_agent(
         user_id=user_id,
-        system_prompt=_FLOATING_SINGLE_AGENT_SYSTEM,
+        system_prompt=system_prompt,
         message=message,
         context=prepared_context,
+        langfuse_prompt=langfuse_prompt,
+        agent_name="floating-agent",
     )
     sanitized = _strip_floating_markup(response)
     if not sanitized and response:
@@ -782,12 +888,17 @@ async def run_home_stream(
     context: dict[str, Any],
 ) -> AsyncGenerator[tuple[str, Any], None]:
     prepared_context = await _prepare_context(message, context)
+    system_prompt, langfuse_prompt = get_prompt_or_fallback(
+        "home_system", _HOME_SYSTEM_PROMPT
+    )
     text_chunks: list[str] = []
     async for event in _run_single_agent_stream(
         user_id=user_id,
-        system_prompt=_HOME_SINGLE_AGENT_SYSTEM,
+        system_prompt=system_prompt,
         message=message,
         context=prepared_context,
+        langfuse_prompt=langfuse_prompt,
+        agent_name="home-agent",
     ):
         event_type, data = event
         if event_type != "token":
@@ -809,14 +920,19 @@ async def run_floating_stream(
     domain = await _infer_floating_domain(message, prepared_context)
     yield "floating_domain", domain
 
+    system_prompt, langfuse_prompt = get_prompt_or_fallback(
+        "floating_system", _FLOATING_SYSTEM_PROMPT
+    )
     sanitizer = _FloatingStreamSanitizer()
     emitted_sanitized = False
     raw_chunks: list[str] = []
     async for event in _run_single_agent_stream(
         user_id=user_id,
-        system_prompt=_FLOATING_SINGLE_AGENT_SYSTEM,
+        system_prompt=system_prompt,
         message=message,
         context=prepared_context,
+        langfuse_prompt=langfuse_prompt,
+        agent_name="floating-agent",
     ):
         event_type, data = event
         if event_type != "token":

app/core/langfuse_client.py

@@ -0,0 +1,147 @@
+"""Langfuse observability — singleton client and prompt helpers.
+
+If LANGFUSE_SECRET_KEY / LANGFUSE_PUBLIC_KEY are not set,
+all helpers are no-ops so the app works without Langfuse configured.
+
+Usage
+-----
+Tracing::
+
+    from app.core.langfuse_client import get_langfuse
+
+    lf = get_langfuse()
+    if lf:
+        with lf.start_as_current_observation(as_type="span", name="my-agent") as span:
+            span.update(input=user_message)
+            # ... do work ...
+            span.update(output=result)
+        lf.flush()
+
+Prompt management::
+
+    from app.core.langfuse_client import get_prompt_or_fallback
+
+    text, prompt_obj = get_prompt_or_fallback("home_system", FALLBACK_PROMPT)
+    # Use text as the system prompt; pass prompt_obj to generations for linking.
+
+Linking a prompt to a generation::
+
+    with lf.start_as_current_observation(
+        as_type="generation",
+        name="llm-call",
+        model="gpt-4o",
+        prompt=prompt_obj,   # links generation → prompt version in the UI
+        input=messages,
+    ) as gen:
+        response = await llm.ainvoke(messages)
+        gen.update(output=response.content, usage=_usage(response))
+"""
+
+from __future__ import annotations
+
+import logging
+from typing import Any
+
+logger = logging.getLogger(__name__)
+
+_client: Any = None
+_initialized: bool = False
+
+
+def get_langfuse() -> Any | None:
+    """Return the Langfuse singleton, or ``None`` when not configured."""
+    global _client, _initialized
+    if _initialized:
+        return _client
+    _initialized = True
+
+    from app.config.settings import settings  # local import to avoid circular deps
+
+    if not settings.LANGFUSE_SECRET_KEY or not settings.LANGFUSE_PUBLIC_KEY:
+        logger.debug("langfuse: not configured — observability disabled")
+        return None
+
+    try:
+        from langfuse import Langfuse
+
+        _client = Langfuse(
+            secret_key=settings.LANGFUSE_SECRET_KEY,
+            public_key=settings.LANGFUSE_PUBLIC_KEY,
+            host=settings.LANGFUSE_HOST,
+        )
+        logger.info("langfuse: client initialized host=%s", settings.LANGFUSE_HOST)
+    except Exception as exc:
+        logger.warning("langfuse: failed to initialize: %s", exc)
+        _client = None
+
+    return _client
+
+
+def get_prompt_or_fallback(name: str, fallback: str) -> tuple[str, Any]:
+    """Fetch a text prompt from Langfuse; fall back to ``fallback`` on any error.
+
+    Returns ``(raw_template, prompt_obj_or_None)``.
+
+    * ``raw_template`` — the uncompiled template string.  Do NOT call ``.format()``
+      on it directly; use :func:`compile_prompt` instead so the correct variable
+      syntax is applied (``{{var}}`` for Langfuse, ``{var}`` for the fallback).
+    * ``prompt_obj`` — the Langfuse prompt object, or ``None`` when Langfuse is
+      unavailable / the fetch failed.  Pass this to generation observations so
+      Langfuse links the generation to the exact prompt version in the UI.
+    """
+    lf = get_langfuse()
+    if lf is None:
+        return fallback, None
+
+    try:
+        prompt = lf.get_prompt(name, label="production", fallback=fallback)
+        # For text-type prompts .prompt holds the raw template string.
+        raw = prompt.prompt if hasattr(prompt, "prompt") and isinstance(prompt.prompt, str) else fallback
+        return raw, prompt
+    except Exception as exc:
+        logger.warning("langfuse: get_prompt %r failed: %s — using fallback", name, exc)
+        return fallback, None
+
+
+def compile_prompt(template: str, prompt_obj: Any, **variables: Any) -> str:
+    """Compile *template* with *variables*, choosing the right syntax.
+
+    * When *prompt_obj* is a real Langfuse prompt object, calls
+      ``prompt_obj.compile(**variables)`` which handles ``{{variable}}``
+      substitution as defined in the Langfuse UI.
+    * When *prompt_obj* is ``None`` (Langfuse unavailable or fetch failed),
+      falls back to ``template.format(**variables)`` which handles the
+      ``{variable}`` syntax used in the hardcoded fallback strings.
+
+    This keeps callers oblivious to which syntax is in use.
+    """
+    if prompt_obj is not None:
+        try:
+            compiled = prompt_obj.compile(**variables)
+            # compile() returns a string for text prompts.
+            if isinstance(compiled, str):
+                return compiled
+            # Chat prompts return a list of dicts — join text parts.
+            if isinstance(compiled, list):
+                return "\n".join(
+                    m.get("content", "") for m in compiled if isinstance(m, dict)
+                )
+        except Exception as exc:
+            logger.warning(
+                "langfuse: compile failed for prompt %r: %s — falling back to .format()",
+                getattr(prompt_obj, "name", "?"),
+                exc,
+            )
+    return template.format(**variables)
+
+
+def extract_usage(response: Any) -> dict[str, int]:
+    """Extract token usage from a LangChain AI message into Langfuse format."""
+    meta = getattr(response, "usage_metadata", None)
+    if not meta:
+        return {}
+    return {
+        "input": int(meta.get("input_tokens", 0)),
+        "output": int(meta.get("output_tokens", 0)),
+        "total": int(meta.get("total_tokens", 0)),
+    }

app/core/preprocessors/__init__.py

@@ -0,0 +1,104 @@
+"""Preprocessor registry: detect content type and dispatch to handlers.
+
+Public API
+----------
+detect_content_type(filename, raw_content) -> str
+    Heuristic detection based on file extension and content patterns.
+
+preprocess(content_type, raw_content) -> PreprocessResult
+    Dispatch to the appropriate handler.
+"""
+
+from __future__ import annotations
+
+import re
+
+from app.core.preprocessors.base import PreprocessResult
+
+# ── Heuristics ────────────────────────────────────────────────────────
+
+# Patterns that strongly suggest an email HTML file
+_EMAIL_SIGNALS = re.compile(
+    r"(Subject:|From:|To:|Date:|Sent:|MIME-Version:|Content-Type:\s*text/html)",
+    re.IGNORECASE,
+)
+
+# Patterns that suggest a generic HTML page (not an email)
+_GENERIC_HTML_SIGNALS = re.compile(
+    r"<(nav|main|header|footer|article|section)\b",
+    re.IGNORECASE,
+)
+
+
+def detect_content_type(filename: str, raw_content: str) -> str:
+    """Return a content-type string for the given file.
+
+    Supported types: ``"email_html"``, ``"generic_html"``,
+    ``"plain_text"``, ``"unknown"``.
+    """
+    ext = filename.rsplit(".", 1)[-1].lower() if "." in filename else ""
+
+    if ext == "txt":
+        return "plain_text"
+
+    if ext in ("html", "htm", "eml", "mhtml", "mht"):
+        # Prefer email detection over generic HTML
+        if _EMAIL_SIGNALS.search(raw_content[:4096]):
+            return "email_html"
+        if _GENERIC_HTML_SIGNALS.search(raw_content[:4096]) or "<html" in raw_content[:200].lower():
+            return "generic_html"
+        # .html without clear signals — check for any email header
+        if re.search(r"^(From|To|Subject|Date):", raw_content[:2048], re.MULTILINE | re.IGNORECASE):
+            return "email_html"
+        return "generic_html"
+
+    # Plain text files with email headers
+    if ext in ("", "txt") or not ext:
+        if _EMAIL_SIGNALS.search(raw_content[:4096]):
+            return "email_html"
+
+    # Detect binary content
+    try:
+        raw_content.encode("utf-8")
+    except (UnicodeEncodeError, AttributeError):
+        return "unknown"
+
+    # Non-text bytes heuristic: high ratio of non-printable chars
+    sample = raw_content[:512]
+    non_printable = sum(1 for c in sample if ord(c) < 32 and c not in "\r\n\t")
+    if len(sample) > 0 and non_printable / len(sample) > 0.1:
+        return "unknown"
+
+    return "unknown"
+
+
+# ── Generic fallback handler ──────────────────────────────────────────
+
+def _preprocess_generic(raw_content: str, content_type: str) -> PreprocessResult:
+    """Strip HTML tags if present, return text as-is."""
+    try:
+        from bs4 import BeautifulSoup
+        text = BeautifulSoup(raw_content, "html.parser").get_text(separator="\n")
+    except ImportError:
+        # No BeautifulSoup — strip tags with a simple regex
+        text = re.sub(r"<[^>]+>", "", raw_content)
+
+    text = re.sub(r"\n{3,}", "\n\n", text).strip()
+    return PreprocessResult(content_type=content_type, clean_text=text, metadata={})
+
+
+# ── Dispatch ──────────────────────────────────────────────────────────
+
+def preprocess(content_type: str, raw_content: str) -> PreprocessResult:
+    """Dispatch *raw_content* to the handler registered for *content_type*.
+
+    Falls back to the generic handler for unknown types.
+    """
+    if content_type == "email_html":
+        from app.core.preprocessors.email_html import preprocess_email_html
+        return preprocess_email_html(raw_content)
+
+    return _preprocess_generic(raw_content, content_type)
+
+
+__all__ = ["detect_content_type", "preprocess", "PreprocessResult"]

app/core/preprocessors/base.py

@@ -0,0 +1,25 @@
+"""Base types for the preprocessor system."""
+
+from __future__ import annotations
+
+from dataclasses import dataclass, field
+
+
+@dataclass
+class PreprocessResult:
+    """Output of a preprocessor handler.
+
+    Attributes
+    ----------
+    content_type:
+        The detected content type (e.g. ``"email_html"``, ``"plain_text"``).
+    clean_text:
+        Human-readable text stripped of markup/binary noise.
+    metadata:
+        Dict of extracted metadata (keys vary by handler).
+        Common keys: ``subject``, ``from``, ``to``, ``date``, ``filename``.
+    """
+
+    content_type: str
+    clean_text: str
+    metadata: dict = field(default_factory=dict)

app/core/preprocessors/email_html.py

@@ -0,0 +1,111 @@
+"""Preprocessor for email HTML files.
+
+Handles:
+- HTML stripping via BeautifulSoup
+- Metadata extraction (Subject, From, To, Date)
+- Thread splitting — isolates the latest reply
+"""
+
+from __future__ import annotations
+
+import re
+from typing import TYPE_CHECKING
+
+from app.core.preprocessors.base import PreprocessResult
+
+if TYPE_CHECKING:
+    pass
+
+# ── Thread split markers ──────────────────────────────────────────────
+
+# Matches patterns like:
+#   "On Mon, Apr 7, 2026 at 10:00 AM, Alice <alice@co.com> wrote:"
+#   "-----Original Message-----"
+#   "> " (plain-text quote prefix)
+_THREAD_PATTERNS = [
+    re.compile(r"^On\s+.+wrote\s*:", re.IGNORECASE | re.MULTILINE),
+    re.compile(r"^-{3,}\s*(original message|forwarded message)\s*-{3,}", re.IGNORECASE | re.MULTILINE),
+    re.compile(r"^>{1,}\s+\S", re.MULTILINE),
+    re.compile(r"^From:\s+.+\nSent:\s+", re.IGNORECASE | re.MULTILINE),
+]
+
+# ── Metadata patterns (applied on raw HTML / plain fallback) ──────────
+
+_META_PATTERNS: dict[str, list[re.Pattern]] = {
+    "subject": [
+        re.compile(r"<title>(.+?)</title>", re.IGNORECASE | re.DOTALL),
+        re.compile(r"Subject:\s*(.+)", re.IGNORECASE),
+    ],
+    "from": [
+        re.compile(r'<meta[^>]+name=["\']?from["\']?[^>]+content=["\']([^"\']+)["\']', re.IGNORECASE),
+        re.compile(r"From:\s*(.+)", re.IGNORECASE),
+    ],
+    "to": [
+        re.compile(r'<meta[^>]+name=["\']?to["\']?[^>]+content=["\']([^"\']+)["\']', re.IGNORECASE),
+        re.compile(r"To:\s*(.+)", re.IGNORECASE),
+    ],
+    "date": [
+        re.compile(r'<meta[^>]+name=["\']?date["\']?[^>]+content=["\']([^"\']+)["\']', re.IGNORECASE),
+        re.compile(r"Date:\s*(.+)", re.IGNORECASE),
+        re.compile(r"Sent:\s*(.+)", re.IGNORECASE),
+    ],
+}
+
+
+def _extract_metadata(raw_html: str, text: str) -> dict:
+    """Extract Subject/From/To/Date from raw HTML or plain text."""
+    metadata: dict[str, str] = {}
+    for field, patterns in _META_PATTERNS.items():
+        for pat in patterns:
+            m = pat.search(raw_html) or pat.search(text)
+            if m:
+                metadata[field] = m.group(1).strip()
+                break
+    return metadata
+
+
+def _split_thread(text: str) -> str:
+    """Return only the latest message in a threaded email."""
+    earliest_pos: int | None = None
+    for pat in _THREAD_PATTERNS:
+        m = pat.search(text)
+        if m and (earliest_pos is None or m.start() < earliest_pos):
+            earliest_pos = m.start()
+
+    if earliest_pos is not None and earliest_pos > 0:
+        return text[:earliest_pos].strip()
+    return text.strip()
+
+
+def preprocess_email_html(raw_content: str) -> PreprocessResult:
+    """Strip HTML, extract metadata, split thread from an email HTML file."""
+    try:
+        from bs4 import BeautifulSoup  # lazy import — optional dep
+    except ImportError as exc:
+        raise ImportError(
+            "beautifulsoup4 is required for email_html preprocessing. "
+            "Install it with: pip install beautifulsoup4"
+        ) from exc
+
+    # Parse with lxml if available, fall back to html.parser
+    try:
+        soup = BeautifulSoup(raw_content, "lxml")
+    except Exception:
+        soup = BeautifulSoup(raw_content, "html.parser")
+
+    # Remove noise tags
+    for tag in soup(["style", "script", "head", "noscript"]):
+        tag.decompose()
+
+    clean_text = soup.get_text(separator="\n")
+    # Collapse excessive blank lines
+    clean_text = re.sub(r"\n{3,}", "\n\n", clean_text).strip()
+
+    metadata = _extract_metadata(raw_content, clean_text)
+    latest_message = _split_thread(clean_text)
+
+    return PreprocessResult(
+        content_type="email_html",
+        clean_text=latest_message,
+        metadata=metadata,
+    )

app/models.py

@@ -296,6 +296,7 @@ class LocalAgentConfig(Base):
     directory_paths: Mapped[list] = mapped_column(JSON, nullable=False, default=list)
     data_types: Mapped[list] = mapped_column(JSON, nullable=False, default=list)
     prompt_template: Mapped[str] = mapped_column(Text, nullable=False, default="")
+    agent_config: Mapped[dict | None] = mapped_column(JSON, nullable=True)
     file_extensions: Mapped[list] = mapped_column(JSON, nullable=False, default=list)
     schedule_cron: Mapped[str] = mapped_column(String(100), nullable=False, default="0 */6 * * *")
     enabled: Mapped[bool] = mapped_column(Boolean, nullable=False, default=True)

app/schemas.py

@@ -273,6 +273,27 @@ class WsFloatingDomain(BaseModel):
     domain: WsDomain
 
 
+# ── Agent Config V2 ───────────────────────────────────────────────────
+
+
+class ContentTypeConfig(BaseModel):
+    """Per-type extraction config produced by the journey chatbot."""
+
+    id: str
+    label: str = ""
+    detection_hint: str = ""
+    preprocessing: str = "generic"  # handler name: "email_html", "plain_text", ...
+    extraction_prompt: str
+
+
+class AgentConfig(BaseModel):
+    """Structured agent configuration (replaces freeform prompt_template)."""
+
+    content_types: list[ContentTypeConfig] = []
+    global_rules: list[str] = []
+    data_types: list[str] = []
+
+
 # ── Agent Catalog ─────────────────────────────────────────────────────
 
 class AgentCatalogItem(BaseModel):

requirements.txt

@@ -32,4 +32,8 @@ google-auth-oauthlib>=1.2.0
 google-auth-httplib2>=0.2.0
 msal>=1.28.0
 cryptography>=42.0.0
+langfuse>=2.0.0
+beautifulsoup4>=4.12.0
+lxml>=5.0.0
+PyYAML>=6.0.0
 ruff>=0.8.0

tests/conftest.py

@@ -6,26 +6,21 @@ a per-test session, and a FastAPI ``TestClient`` wired to use it.
 
 from __future__ import annotations
 
-import json
-import os
 import time
 import uuid
 from collections.abc import AsyncGenerator, Generator
-from unittest.mock import patch
 
-import boto3
 import pytest
 import pytest_asyncio
 from fastapi.testclient import TestClient
 from jose import jwt
-from moto import mock_aws
 from sqlalchemy import StaticPool, event
 from sqlalchemy.ext.asyncio import AsyncSession, async_sessionmaker, create_async_engine
 
 from app.config.settings import settings
 from app.db import Base, get_session
 from app.main import app
-from app.models import Plugin, Subscription, User
+from app.models import Subscription, User
 
 # ── Fixed test user IDs (one per tier) ───────────────────────────────
 
@@ -109,79 +104,6 @@ def client(db_session: AsyncSession) -> Generator[TestClient, None, None]:   # n
     app.dependency_overrides.pop(get_session, None)
 
 
-# ── Seed data helpers ────────────────────────────────────────────────
-
-_SEED_PLUGINS = [
-    Plugin(
-        id="plugin-github-sync",
-        name="GitHub Sync",
-        description="Sync tasks with GitHub Issues and pull requests.",
-        version="1.0.0",
-        author_name="Adiuva",
-        category="productivity",
-        price_cents=0,
-        permissions=json.dumps(["read:tasks", "write:tasks"]),
-        status="approved",
-        s3_package_key="plugins/plugin-github-sync/1.0.0/package.zip",
-        install_count=0,
-        avg_rating=0.0,
-    ),
-    Plugin(
-        id="plugin-slack-notify",
-        name="Slack Notifier",
-        description="Post task and timeline updates to Slack channels.",
-        version="1.2.0",
-        author_name="Adiuva",
-        category="communication",
-        price_cents=499,
-        permissions=json.dumps(["read:tasks", "read:timelines"]),
-        status="approved",
-        s3_package_key="plugins/plugin-slack-notify/1.2.0/package.zip",
-        install_count=0,
-        avg_rating=0.0,
-    ),
-    Plugin(
-        id="plugin-time-tracker",
-        name="Time Tracker",
-        description="Track time spent on tasks with automatic reporting.",
-        version="0.9.1",
-        author_name="Third Party",
-        category="productivity",
-        price_cents=999,
-        permissions=json.dumps(["read:tasks", "write:tasks"]),
-        status="approved",
-        s3_package_key="plugins/plugin-time-tracker/0.9.1/package.zip",
-        install_count=0,
-        avg_rating=0.0,
-    ),
-]
-
-
-@pytest_asyncio.fixture
-async def seed_plugins(db_session: AsyncSession) -> list[Plugin]:
-    """Insert the 3 default approved plugins and return them."""
-    plugins = []
-    for template in _SEED_PLUGINS:
-        p = Plugin(
-            id=template.id,
-            name=template.name,
-            description=template.description,
-            version=template.version,
-            author_name=template.author_name,
-            category=template.category,
-            price_cents=template.price_cents,
-            permissions=template.permissions,
-            status=template.status,
-            s3_package_key=template.s3_package_key,
-            install_count=template.install_count,
-            avg_rating=template.avg_rating,
-        )
-        db_session.add(p)
-        plugins.append(p)
-    await db_session.commit()
-    return plugins
-
-
 # ── JWT helpers ──────────────────────────────────────────────────────
 
 
@@ -212,24 +134,21 @@ def auth_header(tier: str = "power", user_id: str | None = None) -> dict[str, st
     return {"Authorization": f"Bearer {make_jwt(tier, user_id)}"}
 
 
-# ── S3 mock fixture ──────────────────────────────────────────────────
+# ── CLI options ───────────────────────────────────────────────────────
 
-S3_TEST_BUCKET = "test-bucket"
-S3_TEST_REGION = "us-east-1"
-
-
-@pytest.fixture
-def s3_bucket():
-    """Create a mocked S3 bucket via moto and patch BlobStore settings."""
-    with mock_aws():
-        os.environ.setdefault("AWS_ACCESS_KEY_ID", "testing")
-        os.environ.setdefault("AWS_SECRET_ACCESS_KEY", "testing")
-        os.environ.setdefault("AWS_DEFAULT_REGION", S3_TEST_REGION)
-        client = boto3.client("s3", region_name=S3_TEST_REGION)
-        client.create_bucket(Bucket=S3_TEST_BUCKET)
-        with patch("app.storage.blob_store.settings") as mock_settings:
-            mock_settings.S3_BUCKET = S3_TEST_BUCKET
-            mock_settings.S3_REGION = S3_TEST_REGION
-            mock_settings.AWS_ACCESS_KEY_ID = "testing"
-            mock_settings.AWS_SECRET_ACCESS_KEY = "testing"
-            yield S3_TEST_BUCKET
+def pytest_addoption(parser):
+    parser.addoption(
+        "--preprocess-dir",
+        default=None,
+        help="Override fixture folder for preprocessor tests (must contain cases.yaml + data/)",
+    )
+    parser.addoption(
+        "--runner-dir",
+        default=None,
+        help="Override fixture folder for agent_runner_v2 eval tests (must contain cases.yaml + data/)",
+    )
+    parser.addoption(
+        "--journey-dir",
+        default=None,
+        help="Override fixture folder for journey_v2 eval tests (must contain cases.yaml + data/)",
+    )

tests/fixtures/agent_runner_v2/cases.yaml

@@ -0,0 +1,86 @@
+# Agent Runner V2 — eval test cases (Step 2, requires real LLM)
+#
+# Each case drives one parametrized `test_eval_runner` invocation.
+#
+# Keys
+# ----
+# id: str                     unique identifier shown in pytest output
+# description: str            human-readable label
+# file: str                   filename inside data/
+# file_path: str              path reported to the executor (affects project-matching via filename)
+# projects: [alpha|beta]      symbolic project names resolved by the test helper
+#
+# Optional pre-existing records (dedup tests)
+# existing_tasks:             list of {id, title, status, priority}
+# existing_notes:             list of {id, title, content}
+# existing_timelines:         list of {id, title, date}
+#
+# Assertions (one or more)
+# expect_insert: <table>      at least 1 insert row in this table (tasks|notes|timelines)
+# expect_no_insert: true      zero inserts in any table
+# expect_project_id: <id>     any insert must carry this projectId
+# expect_dedup: true          task inserts == 0 OR task updates >= 1 (dedup check)
+#
+# Langfuse
+# score_name: str             observation score name
+
+- id: "2.1"
+  description: "Action email → create_task"
+  file: email_action.html
+  file_path: /emails/ProjectAlpha_action.html
+  projects: [alpha, beta]
+  expect_insert: tasks
+  score_name: runner.email_to_task
+
+- id: "2.2"
+  description: "Informational email → create_note"
+  file: email_info.html
+  file_path: /emails/ProjectAlpha_info.html
+  projects: [alpha, beta]
+  expect_insert: notes
+  score_name: runner.email_to_note
+
+- id: "2.3"
+  description: "Email with meeting date → create_timeline"
+  file: email_date.html
+  file_path: /emails/ProjectAlpha_kickoff.html
+  projects: [alpha, beta]
+  expect_insert: timelines
+  score_name: runner.email_to_timeline
+
+- id: "2.4"
+  description: "Filename contains project name → correct project assigned"
+  file: email_action.html
+  file_path: /emails/ProjectAlpha_report.html
+  projects: [alpha, beta]
+  expect_project_id: proj-alpha
+  score_name: runner.project_filename
+
+- id: "2.5"
+  description: "Email body mentions project → correct project assigned"
+  file: email_action.html
+  file_path: /emails/email_001.html
+  projects: [alpha, beta]
+  expect_project_id: proj-alpha
+  score_name: runner.project_content
+
+- id: "2.6"
+  description: "Newsletter + global rule no-project → no creates"
+  file: email_no_project.html
+  file_path: /emails/newsletter.html
+  projects: [alpha, beta]
+  expect_no_insert: true
+  score_name: runner.no_project
+
+- id: "2.7"
+  description: "Existing task with same title → dedup (update not create)"
+  file: email_action.html
+  file_path: /emails/ProjectAlpha_followup.html
+  projects: [alpha]
+  existing_tasks:
+    - id: task-existing
+      title: Fix the login bug
+      status: todo
+      priority: medium
+  expect_dedup: true
+  score_name: runner.dedup

tests/fixtures/agent_runner_v2/data/email_action.html

@@ -0,0 +1,7 @@
+<html><head></head><body>
+<p><b>From:</b> boss@company.com</p>
+<p><b>To:</b> dev@company.com</p>
+<p><b>Subject:</b> Fix the login bug</p>
+<p><b>Date:</b> 2026-04-07</p>
+<p>Hi,<br>Please fix the login bug in Project Alpha by Friday. High priority!</p>
+</body></html>

tests/fixtures/agent_runner_v2/data/email_date.html

@@ -0,0 +1,5 @@
+<html><head></head><body>
+<p><b>From:</b> pm@company.com</p>
+<p><b>Subject:</b> Project Alpha kick-off meeting</p>
+<p>The kick-off meeting for Project Alpha is scheduled for 2026-04-15 at 10:00.</p>
+</body></html>

tests/fixtures/agent_runner_v2/data/email_info.html

@@ -0,0 +1,7 @@
+<html><head></head><body>
+<p><b>From:</b> pm@company.com</p>
+<p><b>To:</b> team@company.com</p>
+<p><b>Subject:</b> FYI: New policy for Project Alpha</p>
+<p>Just a heads-up that starting next week all code reviews must be done
+within 24 hours for Project Alpha. No action needed from you now.</p>
+</body></html>

tests/fixtures/agent_runner_v2/data/email_no_project.html

@@ -0,0 +1,5 @@
+<html><head></head><body>
+<p><b>From:</b> newsletter@ads.com</p>
+<p><b>Subject:</b> Weekly newsletter</p>
+<p>Check out our latest deals on electronics!</p>
+</body></html>

tests/fixtures/journey_v2/cases.yaml

@@ -0,0 +1,87 @@
+# Journey V2 eval test cases — Step 4
+#
+# Each case simulates a complete journey session:
+#   1. handle_journey_start is called with directory + data_types
+#   2. handle_journey_message is called for each entry in user_messages
+#   3. Assertions are evaluated on the final reply
+#
+# directory_files: list of {path, content_file} — content_file is relative to data/
+#
+# Assertion keys:
+#   expect_question: true          → first reply must contain "?"
+#   expect_done: true              → final reply must have done=True
+#   expect_valid_config: true      → agent_config must be parseable as AgentConfig with content_types > 0
+#   expect_content_type_id: <str>  → AgentConfig.content_types must contain an entry with this id
+#   expect_extraction_contains: <str> → first content_type extraction_prompt must contain this word
+#   expect_global_rules: true      → AgentConfig.global_rules must be non-empty
+
+- id: "4.1"
+  description: "Journey start explores directory, first reply contains a question"
+  directory: "/test/emails"
+  data_types: ["tasks", "notes", "timelines"]
+  directory_files:
+    - path: "/test/emails/outlook_export_2024.html"
+      content_file: "email_action.html"
+  user_messages: []
+  score_name: "journey.start"
+  expect_question: true
+
+- id: "4.2"
+  description: "Full 3-turn conversation produces a valid AgentConfig JSON"
+  directory: "/test/emails"
+  data_types: ["tasks", "notes", "timelines"]
+  directory_files:
+    - path: "/test/emails/email_backup.html"
+      content_file: "email_action.html"
+  user_messages:
+    - "These are email exports from Outlook in HTML format"
+    - "Create tasks for emails with direct action requests, notes for informational emails"
+    - "Yes, that looks correct. No other rules."
+  score_name: "journey.valid_json"
+  expect_done: true
+  expect_valid_config: true
+
+- id: "4.3"
+  description: "Journey detects email_html content type from directory exploration"
+  directory: "/test/emails"
+  data_types: ["tasks", "notes"]
+  directory_files:
+    - path: "/test/emails/message.html"
+      content_file: "email_action.html"
+  user_messages:
+    - "HTML email backups from my mail client, exported from Outlook"
+    - "Create tasks from emails that contain assignments or direct action items"
+    - "Correct, no other rules needed"
+  score_name: "journey.detect_email"
+  expect_done: true
+  expect_content_type_id: "email_html"
+
+- id: "4.4"
+  description: "Custom user rule (only notes, no tasks) reflected in extraction_prompt"
+  directory: "/test/emails"
+  data_types: ["notes"]
+  directory_files:
+    - path: "/test/emails/email.html"
+      content_file: "email_info.html"
+  user_messages:
+    - "HTML emails from my work inbox"
+    - "Create only notes from all emails — I do not want tasks or timelines to be created"
+    - "Yes, exactly"
+  score_name: "journey.custom_rules"
+  expect_done: true
+  expect_extraction_contains: "note"
+
+- id: "4.5"
+  description: "Global rule (no project = no entity) appears in AgentConfig.global_rules"
+  directory: "/test/emails"
+  data_types: ["tasks", "notes"]
+  directory_files:
+    - path: "/test/emails/email.html"
+      content_file: "email_action.html"
+  user_messages:
+    - "Email backups from Outlook"
+    - "Create tasks from action request emails, notes from informational emails"
+    - "If the email cannot be matched to any project, do not create any entity at all"
+  score_name: "journey.global_rules"
+  expect_done: true
+  expect_global_rules: true

tests/fixtures/journey_v2/data/email_action.html

@@ -0,0 +1,23 @@
+<!DOCTYPE html>
+<html>
+<head>
+  <meta charset="UTF-8">
+  <title>Email: Fix the login bug</title>
+  <style>body { font-family: Arial; } .header { color: #666; }</style>
+</head>
+<body>
+  <div class="header">
+    <p><strong>From:</strong> boss@company.com</p>
+    <p><strong>To:</strong> dev@company.com</p>
+    <p><strong>Subject:</strong> Fix the login bug</p>
+    <p><strong>Date:</strong> Mon, 7 Apr 2026 09:15:00 +0000</p>
+  </div>
+  <div class="body">
+    <p>Hi,</p>
+    <p>Please fix the login bug in Project Alpha as soon as possible.
+    Users are reporting that they can't log in with their Google accounts.
+    This is blocking the whole team. Please resolve it by Friday.</p>
+    <p>Thanks,<br>Boss</p>
+  </div>
+</body>
+</html>

tests/fixtures/journey_v2/data/email_info.html

@@ -0,0 +1,23 @@
+<!DOCTYPE html>
+<html>
+<head>
+  <meta charset="UTF-8">
+  <title>Email: New policy update</title>
+  <style>body { font-family: Arial; }</style>
+</head>
+<body>
+  <div class="header">
+    <p><strong>From:</strong> hr@company.com</p>
+    <p><strong>To:</strong> all@company.com</p>
+    <p><strong>Subject:</strong> FYI: New remote work policy effective May 1</p>
+    <p><strong>Date:</strong> Tue, 8 Apr 2026 10:00:00 +0000</p>
+  </div>
+  <div class="body">
+    <p>Hi everyone,</p>
+    <p>Just a heads-up that starting May 1, 2026 the company will be moving to
+    a hybrid work model. You will be expected to come into the office at least
+    two days per week. More details will follow in the employee handbook.</p>
+    <p>Best,<br>HR Team</p>
+  </div>
+</body>
+</html>

tests/fixtures/preprocessors/cases.yaml

@@ -0,0 +1,68 @@
+# Preprocessor test cases
+#
+# detect: <expected_type>   → chiama detect_content_type(filename, content)
+# process: <content_type>   → chiama preprocess(content_type, content)
+#
+# Sorgente: file: <nome in data/>  oppure  generate: binary_noise
+#
+# Assertions piatte (solo per process):
+#   no_html: true           clean_text senza tag HTML
+#   min_chars: N            len(clean_text) >= N
+#   ratio_lt: F             len(clean) / len(raw) < F
+#   has_meta: [k, ...]      chiavi presenti in metadata
+#   contains: str | [str]   substring(s) presenti in clean_text
+#   excludes: str | [str]   substring(s) assenti da clean_text
+#   content_type: str       result.content_type == questo valore
+
+- id: "1.1"
+  file: email_action.html
+  detect: email_html
+
+- id: "1.2"
+  file: generic_page.html
+  detect: generic_html
+
+- id: "1.3"
+  file: notes.txt
+  detect: plain_text
+
+- id: "1.4"
+  file: archive.xyz
+  generate: binary_noise
+  detect: unknown
+
+- id: "1.5"
+  file: email_action.html
+  process: email_html
+  no_html: true
+  min_chars: 50
+  ratio_lt: 0.8
+
+- id: "1.6"
+  file: email_action.html
+  process: email_html
+  has_meta: [subject, from]
+
+- id: "1.7"
+  file: email_thread.html
+  process: email_html
+  contains: "Sure, I'll handle the deploy"
+  excludes: "Let's plan the deploy"
+
+- id: "1.8"
+  file: email_single.html
+  process: email_html
+  contains: "deploy is done"
+
+- id: "1.9"
+  file: email_heavy.html
+  process: email_html
+  no_html: true
+  min_chars: 30
+  excludes: [border-collapse, font-size]
+
+- id: "1.10"
+  file: fallback.txt
+  process: unknown
+  min_chars: 1
+  content_type: unknown

tests/fixtures/preprocessors/data/email_action.html

@@ -0,0 +1,25 @@
+<!DOCTYPE html>
+<html>
+<head>
+  <title>Fix the login bug</title>
+  <style>
+    body { font-family: Arial, sans-serif; color: #333; margin: 0; padding: 20px; }
+    .header { background: #f5f5f5; padding: 10px; border-bottom: 1px solid #ddd; }
+    .body { padding: 20px; }
+  </style>
+</head>
+<body>
+  <div class="header">
+    <p><strong>From:</strong> boss@company.com</p>
+    <p><strong>To:</strong> dev@company.com</p>
+    <p><strong>Subject:</strong> Fix the login bug</p>
+    <p><strong>Date:</strong> Mon, 7 Apr 2026 09:00:00 +0200</p>
+  </div>
+  <div class="body">
+    <p>Hi,</p>
+    <p>Please fix the login bug by Friday. It is blocking the release.</p>
+    <p>Priority: high. Let me know if you need anything.</p>
+    <p>Thanks,<br>Boss</p>
+  </div>
+</body>
+</html>

tests/fixtures/preprocessors/data/email_heavy.html

@@ -0,0 +1,49 @@
+<!DOCTYPE html>
+<html>
+<head>
+<style>
+  table { border-collapse: collapse; width: 100%; max-width: 600px; margin: 0 auto; }
+  td { padding: 8px 12px; border: 1px solid #dddddd; font-size: 12px; color: #444444; }
+  .header-row { background-color: #003366; color: #ffffff; font-weight: bold; }
+  .label-col { background-color: #f0f0f0; width: 80px; font-weight: bold; }
+  .footer-row { font-size: 10px; color: #999999; text-align: center; }
+</style>
+</head>
+<body bgcolor="#eeeeee">
+<center>
+<table cellpadding="0" cellspacing="0">
+  <tr class="header-row">
+    <td colspan="2">Company Internal Update</td>
+  </tr>
+  <tr>
+    <td class="label-col">From:</td>
+    <td>newsletter@corp.com</td>
+  </tr>
+  <tr>
+    <td class="label-col">Subject:</td>
+    <td>Q1 Results Update</td>
+  </tr>
+  <tr>
+    <td class="label-col">Date:</td>
+    <td>Apr 7, 2026</td>
+  </tr>
+  <tr>
+    <td colspan="2">
+      <table width="100%" cellpadding="10">
+        <tr>
+          <td>
+            <p style="font-size:14px; font-weight:bold;">Dear Team,</p>
+            <p>Q1 results are in. Revenue up 15% year-over-year.</p>
+            <p>Please review the attached report and share any feedback by EOW.</p>
+          </td>
+        </tr>
+      </table>
+    </td>
+  </tr>
+  <tr class="footer-row">
+    <td colspan="2">Confidential — do not forward outside the company.</td>
+  </tr>
+</table>
+</center>
+</body>
+</html>

tests/fixtures/preprocessors/data/email_single.html

@@ -0,0 +1,8 @@
+<!DOCTYPE html>
+<html><body>
+  <p><strong>From:</strong> alice@co.com</p>
+  <p><strong>To:</strong> team@co.com</p>
+  <p><strong>Subject:</strong> Quick update</p>
+  <p><strong>Date:</strong> Tue, 7 Apr 2026 10:30:00 +0200</p>
+  <p>The deploy is done. Everything looks good. No issues so far.</p>
+</body></html>

tests/fixtures/preprocessors/data/email_thread.html

@@ -0,0 +1,24 @@
+<!DOCTYPE html>
+<html><body>
+  <div class="message-latest">
+    <p><strong>From:</strong> alice@co.com</p>
+    <p><strong>Subject:</strong> Re: Re: Deploy plan</p>
+    <p>Sure, I'll handle the deploy.</p>
+  </div>
+
+  <p>On Mon, Apr 6, 2026 at 3:00 PM, Bob &lt;bob@co.com&gt; wrote:</p>
+  <blockquote>
+    <p>From: bob@co.com</p>
+    <p>Can you handle the deploy?</p>
+    <p>On Sun, Apr 5, 2026 at 1:00 PM, Alice &lt;alice@co.com&gt; wrote:</p>
+    <blockquote>
+      <p>From: alice@co.com</p>
+      <p>Let's plan the deploy for Monday.</p>
+      <p>On Sat, Apr 4, 2026 at 11:00 AM, Charlie &lt;charlie@co.com&gt; wrote:</p>
+      <blockquote>
+        <p>From: charlie@co.com</p>
+        <p>We need to schedule the deploy. What day works?</p>
+      </blockquote>
+    </blockquote>
+  </blockquote>
+</body></html>

tests/fixtures/preprocessors/data/fallback.txt

@@ -0,0 +1,3 @@
+random text content without any structure
+line two with some words
+line three and more content here

tests/fixtures/preprocessors/data/generic_page.html

@@ -0,0 +1,35 @@
+<!DOCTYPE html>
+<html lang="en">
+<head>
+  <meta charset="UTF-8">
+  <title>My Web App</title>
+  <link rel="stylesheet" href="styles.css">
+</head>
+<body>
+  <nav>
+    <a href="/">Home</a>
+    <a href="/about">About</a>
+    <a href="/contact">Contact</a>
+  </nav>
+  <main>
+    <header>
+      <h1>Welcome to My App</h1>
+    </header>
+    <article>
+      <p>This is a generic web page with no email headers.</p>
+      <p>It has navigation, main content, and a footer.</p>
+    </article>
+    <section>
+      <h2>Features</h2>
+      <ul>
+        <li>Fast</li>
+        <li>Reliable</li>
+        <li>Secure</li>
+      </ul>
+    </section>
+  </main>
+  <footer>
+    <p>&copy; 2026 My App</p>
+  </footer>
+</body>
+</html>

tests/fixtures/preprocessors/data/notes.txt

@@ -0,0 +1,15 @@
+Meeting notes - April 7, 2026
+
+Attendees: Alice, Bob, Charlie
+
+Discussion points:
+- Deploy scheduled for Friday
+- Bug fix for login must be completed by Thursday
+- Review Q1 numbers before EOW
+
+Action items:
+- Alice: fix login bug
+- Bob: prepare deploy checklist
+- Charlie: send Q1 report
+
+Next meeting: April 14, 2026

tests/test_agent_runner_v2.py

@@ -0,0 +1,432 @@
+"""Tests for Local Agent V2 runner (Step 2).
+
+Covers the unified per-file flow:
+  Phase A — detect + preprocess (Python, zero LLM)
+  Phase B — single LLM call with tools (classify + extract + create)
+
+Fixture-based eval tests (2.1–2.7)
+-----------------------------------
+Cases are defined in tests/fixtures/agent_runner_v2/cases.yaml.
+Email HTML files live in tests/fixtures/agent_runner_v2/data/.
+Use --runner-dir to point at a custom folder (same structure required).
+
+Unit tests (no LLM)
+--------------------
+  2.8  items_created count   → items_created == N create_* calls
+  2.9  Device offline        → status=error
+  2.10 Empty file            → items_processed=0, status=success
+
+Run:
+    pytest tests/test_agent_runner_v2.py -v
+    pytest tests/test_agent_runner_v2.py -v -k "2_9 or 2_10 or 2_8"   # unit only
+    pytest tests/test_agent_runner_v2.py -v -k "eval"                  # LLM evals only
+    pytest tests/test_agent_runner_v2.py -v --runner-dir /path/to/dir  # custom fixtures
+"""
+
+from __future__ import annotations
+
+import uuid
+from contextlib import nullcontext
+from datetime import datetime, timezone
+from pathlib import Path
+from typing import Any
+from unittest.mock import AsyncMock, MagicMock, patch
+
+import pytest
+import yaml
+
+from app.core.agent_runner import (
+    _format_metadata,
+    _format_projects,
+    _get_extraction_rules,
+    _get_no_match_behavior,
+    _is_overdue,
+    run_local_agent,
+)
+from app.core.device_manager import DeviceConnectionManager
+from app.core.langfuse_client import get_langfuse
+from app.models import AgentRunLog, LocalAgentConfig
+from tests.conftest import TEST_USER_IDS
+
+# ── Constants ─────────────────────────────────────────────────────────────
+
+_USER_ID = TEST_USER_IDS["power"]
+
+_DEFAULT_FIXTURE_DIR = Path(__file__).parent / "fixtures" / "agent_runner_v2"
+
+_AGENT_CONFIG = {
+    "content_types": [
+        {
+            "id": "email_html",
+            "label": "Email HTML",
+            "detection_hint": "HTML file with From/To/Subject headers",
+            "preprocessing": "email_html",
+            "extraction_prompt": (
+                "If the email contains a direct action request or task assignment → create a task. "
+                "If the email contains informational content, updates, or FYI → create a note. "
+                "If the email mentions a specific date for a meeting or deadline → create a timeline entry."
+            ),
+        }
+    ],
+    "global_rules": [
+        "Se il file non è riconducibile a nessun progetto, non creare alcuna entità."
+    ],
+    "data_types": ["tasks", "notes", "timelines"],
+}
+
+# Canonical project definitions, referenced symbolically in cases.yaml.
+_PROJECTS: dict[str, dict] = {
+    "alpha": {"id": "proj-alpha", "name": "Project Alpha", "status": "active"},
+    "beta":  {"id": "proj-beta",  "name": "Project Beta",  "status": "active"},
+}
+
+
+# ── Fixture loading ───────────────────────────────────────────────────────
+
+
+def _fixtures_dir(config) -> Path:
+    override = config.getoption("--runner-dir")
+    return Path(override) if override else _DEFAULT_FIXTURE_DIR
+
+
+def _load_cases(config) -> list[dict]:
+    return yaml.safe_load(
+        (_fixtures_dir(config) / "cases.yaml").read_text(encoding="utf-8")
+    )
+
+
+def _read_case_file(case: dict, data_dir: Path) -> str:
+    return (data_dir / case["file"]).read_text(encoding="utf-8")
+
+
+def _resolve_projects(entries: list[str | dict]) -> list[dict]:
+    """Resolve project list from YAML: symbolic names and/or inline dicts."""
+    result = []
+    for entry in entries:
+        if isinstance(entry, str):
+            if entry in _PROJECTS:
+                result.append(_PROJECTS[entry])
+        elif isinstance(entry, dict):
+            result.append(entry)
+    return result
+
+
+# ── pytest_generate_tests — parametrize eval tests from YAML ─────────────
+
+
+def pytest_generate_tests(metafunc):
+    if "runner_case" not in metafunc.fixturenames:
+        return
+    cases = _load_cases(metafunc.config)
+    metafunc.parametrize("runner_case", cases, ids=[c["id"] for c in cases])
+
+
+# ── Test helpers ──────────────────────────────────────────────────────────
+
+
+def _make_config(
+    agent_config: dict | None = None,
+    directory: str = "/emails",
+    device_id: str = "dev-001",
+) -> LocalAgentConfig:
+    return LocalAgentConfig(
+        id=str(uuid.uuid4()),
+        user_id=_USER_ID,
+        device_id=device_id,
+        name="Test V2 Agent",
+        directory_paths=[directory],
+        data_types=["tasks", "notes", "timelines"],
+        prompt_template="",
+        agent_config=agent_config or _AGENT_CONFIG,
+        file_extensions=[".html", ".eml"],
+        schedule_cron="0 */6 * * *",
+        enabled=True,
+        last_run_at=None,
+    )
+
+
+def _make_run_log(agent_id: str) -> AgentRunLog:
+    return AgentRunLog(
+        id=str(uuid.uuid4()),
+        agent_id=agent_id,
+        agent_type="local",
+        user_id=_USER_ID,
+        status="running",
+        started_at=datetime.now(timezone.utc),
+    )
+
+
+def _make_manager(online: bool = True) -> DeviceConnectionManager:
+    mgr = DeviceConnectionManager()
+    if online:
+        ws = MagicMock()
+        ws.send_text = AsyncMock()
+        mgr.register(_USER_ID, "dev-001", ws)
+    return mgr
+
+
+def _make_executor(
+    file_path: str,
+    file_content: str,
+    projects: list[dict] | None = None,
+    existing_tasks: list[dict] | None = None,
+    existing_notes: list[dict] | None = None,
+    existing_timelines: list[dict] | None = None,
+) -> tuple[Any, list[dict]]:
+    """Return (async_executor, captured_calls).
+
+    The executor handles all ``execute_on_client`` payloads:
+    directory listing, file reading, project/entity fetching, and CRUD.
+    """
+    calls: list[dict] = []
+    _projects = projects if projects is not None else list(_PROJECTS.values())
+
+    async def _executor(payload: dict) -> dict:
+        action = payload.get("action", "")
+        table = payload.get("table", "")
+        data = payload.get("data") or {}
+        calls.append({"action": action, "table": table, "data": data})
+
+        if action == "list_directory":
+            return {"entries": [{"type": "file", "path": file_path}]}
+
+        if action == "get_file_metadata":
+            return {"modifiedAt": None}
+
+        if action == "read_file_content":
+            return {"content": file_content}
+
+        if action == "select":
+            if table == "projects":
+                return {"rows": _projects}
+            if table == "tasks":
+                return {"rows": existing_tasks or []}
+            if table == "notes":
+                return {"rows": existing_notes or []}
+            if table == "timelines":
+                return {"rows": existing_timelines or []}
+            return {"rows": []}
+
+        if action == "insert":
+            return {"row": {"id": str(uuid.uuid4()), **data}}
+
+        if action == "update":
+            return {"success": True}
+
+        return {}
+
+    return _executor, calls
+
+
+# ── Unit: helper functions ────────────────────────────────────────────────
+
+
+def test_format_projects_empty():
+    assert "(no projects" in _format_projects([])
+
+
+def test_format_projects_with_data():
+    result = _format_projects([_PROJECTS["alpha"]])
+    assert "proj-alpha" in result
+    assert "Project Alpha" in result
+
+
+def test_format_metadata_empty():
+    assert _format_metadata({}) == ""
+
+
+def test_format_metadata_email():
+    meta = {"subject": "Fix bug", "from": "boss@co.com", "date": "2026-04-07"}
+    result = _format_metadata(meta)
+    assert "Fix bug" in result
+    assert "boss@co.com" in result
+
+
+def test_get_extraction_rules_match():
+    rules = _get_extraction_rules(_AGENT_CONFIG, "email_html")
+    assert "task" in rules.lower()
+
+
+def test_get_extraction_rules_fallback():
+    rules = _get_extraction_rules(_AGENT_CONFIG, "plain_text")
+    assert "extract" in rules.lower()
+
+
+def test_get_no_match_behavior_from_global_rules():
+    behavior = _get_no_match_behavior(_AGENT_CONFIG)
+    assert behavior  # non-empty
+
+
+def test_get_no_match_behavior_default():
+    behavior = _get_no_match_behavior({})
+    assert "project" in behavior.lower()
+
+
+# ── Unit: 2.9 — device offline ───────────────────────────────────────────
+
+
+@pytest.mark.asyncio
+async def test_2_9_device_offline():
+    """2.9 No device online → status=error, no executor created."""
+    config = _make_config()
+    run_log = _make_run_log(config.id)
+    mgr = _make_manager(online=False)
+
+    with patch("app.core.agent_runner._finalize_run", new_callable=AsyncMock) as mock_fin:
+        await run_local_agent(_USER_ID, config, run_log, mgr)
+
+    _, kwargs = mock_fin.call_args
+    assert kwargs["status"] == "error"
+    assert any("not connected" in e for e in kwargs.get("errors", []))
+
+
+# ── Unit: 2.10 — empty file ──────────────────────────────────────────────
+
+
+@pytest.mark.asyncio
+async def test_2_10_empty_file():
+    """2.10 File with empty content → skipped, items_processed=0, success."""
+    config = _make_config()
+    run_log = _make_run_log(config.id)
+    mgr = _make_manager()
+
+    executor, calls = _make_executor(
+        file_path="/emails/empty.html",
+        file_content="",
+        projects=[_PROJECTS["alpha"]],
+    )
+
+    with patch("app.core.agent_runner._make_agent_executor", return_value=executor), \
+         patch("app.core.agent_runner._finalize_run", new_callable=AsyncMock) as mock_fin:
+        await run_local_agent(_USER_ID, config, run_log, mgr)
+
+    _, kwargs = mock_fin.call_args
+    assert kwargs["items_processed"] == 0
+    assert kwargs["status"] == "success"
+    assert kwargs["items_created"] == 0
+
+
+# ── Unit: 2.8 — items_created count ─────────────────────────────────────
+
+
+@pytest.mark.asyncio
+async def test_2_8_items_created_count():
+    """2.8 items_created == number of create_* tool calls per run."""
+    config = _make_config()
+    run_log = _make_run_log(config.id)
+    mgr = _make_manager()
+
+    executor, _calls = _make_executor(
+        file_path="/emails/action.html",
+        file_content="<html><body><p>Fix the login bug in Project Alpha.</p></body></html>",
+        projects=[_PROJECTS["alpha"]],
+    )
+
+    async def mock_run_agent(*, _tool_calls_out=None, **kw) -> str:
+        if _tool_calls_out is not None:
+            _tool_calls_out.extend(["create_task", "create_note", "update_task"])
+        return "Done."
+
+    with patch("app.core.agent_runner._make_agent_executor", return_value=executor), \
+         patch("app.core.agent_runner._run_agent_with_tools", side_effect=mock_run_agent), \
+         patch("app.core.agent_runner._finalize_run", new_callable=AsyncMock) as mock_fin:
+        await run_local_agent(_USER_ID, config, run_log, mgr)
+
+    _, kwargs = mock_fin.call_args
+    # Only create_task + create_note count (not update_task).
+    assert kwargs["items_created"] == 2
+    assert kwargs["items_processed"] == 1
+
+
+# ── Eval: 2.1–2.7 — fixture-driven, real LLM + Langfuse scoring ──────────
+#
+# Cases loaded from tests/fixtures/agent_runner_v2/cases.yaml.
+# Supported assertions (from YAML):
+#   expect_insert: <table>   → at least 1 insert in that table
+#   expect_no_insert: true   → zero inserts in any table
+#   expect_project_id: <id>  → any insert carries this projectId
+#   expect_dedup: true       → task inserts == 0 OR task updates >= 1
+# ─────────────────────────────────────────────────────────────────────────
+
+
+@pytest.mark.asyncio
+@pytest.mark.eval
+async def test_eval_runner(runner_case, pytestconfig):
+    """Parametrized eval test — one invocation per YAML case."""
+    case: dict = runner_case
+    data_dir = _fixtures_dir(pytestconfig) / "data"
+    file_content = _read_case_file(case, data_dir)
+    projects = _resolve_projects(case.get("projects", []))
+
+    config = _make_config()
+    run_log = _make_run_log(config.id)
+    mgr = _make_manager()
+
+    executor, calls = _make_executor(
+        file_path=case["file_path"],
+        file_content=file_content,
+        projects=projects,
+        existing_tasks=case.get("existing_tasks"),
+        existing_notes=case.get("existing_notes"),
+        existing_timelines=case.get("existing_timelines"),
+    )
+
+    lf = get_langfuse()
+    obs_ctx = lf.start_as_current_observation(
+        name=f"eval-runner-{case['id']}-{case.get('score_name', 'unknown').replace('.', '-')}",
+        metadata={"step": "2", "case_id": case["id"]},
+    ) if lf else nullcontext()
+
+    with obs_ctx as obs:
+        with patch("app.core.agent_runner._make_agent_executor", return_value=executor), \
+             patch("app.core.agent_runner._finalize_run", new_callable=AsyncMock) as mock_fin:
+            await run_local_agent(_USER_ID, config, run_log, mgr)
+
+        _, kwargs = mock_fin.call_args
+        inserts = [c for c in calls if c["action"] == "insert"]
+        score, comment = _evaluate_case(case, calls, kwargs)
+
+        if obs is not None:
+            obs.score(
+                name=case.get("score_name", f"runner.case_{case['id']}"),
+                value=score,
+                comment=comment,
+            )
+
+    if lf:
+        lf.flush()
+
+    assert score == 1.0, f"[{case['id']}] {case.get('description', '')} — {comment}"
+
+
+def _evaluate_case(case: dict, calls: list[dict], finalize_kwargs: dict) -> tuple[float, str]:
+    """Return (score, comment) for a YAML case given the captured executor calls."""
+    inserts = [c for c in calls if c["action"] == "insert"]
+
+    if case.get("expect_no_insert"):
+        score = 1.0 if len(inserts) == 0 else 0.0
+        return score, f"inserts={len(inserts)} (expected 0)"
+
+    if "expect_insert" in case:
+        tables = case["expect_insert"]
+        if isinstance(tables, str):
+            tables = [tables]
+        missing = [t for t in tables if not any(c["table"] == t for c in inserts)]
+        score = 1.0 if not missing else 0.0
+        counts = {t: sum(1 for c in inserts if c["table"] == t) for t in tables}
+        return score, f"inserts={counts}" + (f" missing={missing}" if missing else "")
+
+    if "expect_project_id" in case:
+        expected_pid = case["expect_project_id"]
+        correct = any(c.get("data", {}).get("projectId") == expected_pid for c in inserts)
+        score = 1.0 if correct else 0.0
+        all_pids = [c.get("data", {}).get("projectId") for c in inserts]
+        return score, f"projectIds={all_pids} (expected {expected_pid!r})"
+
+    if case.get("expect_dedup"):
+        task_creates = [c for c in inserts if c["table"] == "tasks"]
+        task_updates = [c for c in calls if c["action"] == "update" and c["table"] == "tasks"]
+        score = 1.0 if len(task_creates) == 0 or len(task_updates) >= 1 else 0.0
+        return score, f"task_creates={len(task_creates)} task_updates={len(task_updates)}"
+
+    return 0.0, "no assertion defined in case"

tests/test_journey_v2.py

@@ -0,0 +1,349 @@
+"""Tests for Local Agent V2 journey setup (Step 4).
+
+Covers the chatbot journey that produces a structured AgentConfig JSON
+instead of a freeform prompt_template string.
+
+Unit tests (no LLM)
+--------------------
+  4.6a  _extract_agent_config: valid JSON → returns serialised config
+  4.6b  _extract_agent_config: invalid JSON → returns None
+  4.6c  _extract_agent_config: markers absent → returns None
+  4.6d  _extract_agent_config: only START marker → returns None
+  4.6e  Session not found → done=True, agent_config=None
+  4.6f  Nudge uses AGENT_CONFIG_START/END markers (not old PROMPT_TEMPLATE)
+
+Eval tests (real LLM + Langfuse scoring)
+-----------------------------------------
+Cases are defined in tests/fixtures/journey_v2/cases.yaml.
+Email HTML files live in tests/fixtures/journey_v2/data/.
+Use --journey-dir to point at a custom folder (same structure required).
+
+Run:
+    pytest tests/test_journey_v2.py -v
+    pytest tests/test_journey_v2.py -v -k "4_6"          # unit only
+    pytest tests/test_journey_v2.py -v -k "eval"          # LLM evals only
+    pytest tests/test_journey_v2.py -v --journey-dir /p   # custom fixtures
+"""
+
+from __future__ import annotations
+
+import uuid
+from contextlib import nullcontext
+from pathlib import Path
+from typing import Any
+from unittest.mock import patch
+
+import pytest
+import yaml
+
+from app.api.routes.agent_setup import (
+    _CONFIG_END,
+    _CONFIG_START,
+    _MAX_TURNS,
+    _extract_agent_config,
+    _sessions,
+    handle_journey_message,
+    handle_journey_start,
+)
+from app.core.langfuse_client import get_langfuse
+from app.core.ws_context import clear_client_executor, set_client_executor
+from app.schemas import AgentConfig
+from tests.conftest import TEST_USER_IDS
+
+# ── Constants ─────────────────────────────────────────────────────────────
+
+_USER_ID = TEST_USER_IDS["power"]
+
+_DEFAULT_FIXTURE_DIR = Path(__file__).parent / "fixtures" / "journey_v2"
+
+# ── Fixture loading ───────────────────────────────────────────────────────
+
+
+def _fixtures_dir(config) -> Path:
+    override = config.getoption("--journey-dir")
+    return Path(override) if override else _DEFAULT_FIXTURE_DIR
+
+
+def _load_cases(config) -> list[dict]:
+    return yaml.safe_load(
+        (_fixtures_dir(config) / "cases.yaml").read_text(encoding="utf-8")
+    )
+
+
+def _read_data_file(filename: str, fixtures_dir: Path) -> str:
+    return (fixtures_dir / "data" / filename).read_text(encoding="utf-8")
+
+
+# ── pytest_generate_tests ─────────────────────────────────────────────────
+
+
+def pytest_generate_tests(metafunc):
+    if "journey_case" not in metafunc.fixturenames:
+        return
+    cases = _load_cases(metafunc.config)
+    metafunc.parametrize("journey_case", cases, ids=[c["id"] for c in cases])
+
+
+# ── Executor builder ──────────────────────────────────────────────────────
+
+
+def _make_fs_executor(directory_files: list[dict], fixtures_dir: Path):
+    """Return an async callback that simulates filesystem tool responses.
+
+    Matches the signature expected by ``set_client_executor`` / ``execute_on_client``:
+    receives the full ``payload`` dict and returns a result dict.
+
+    ``directory_files`` is a list of ``{path, content_file}`` dicts;
+    ``content_file`` is relative to ``fixtures_dir/data/``.
+    """
+    file_map: dict[str, str] = {
+        entry["path"]: _read_data_file(entry["content_file"], fixtures_dir)
+        for entry in directory_files
+    }
+
+    async def _executor(payload: dict) -> dict:
+        action = payload.get("action", "")
+        data = payload.get("data") or {}
+
+        if action == "list_directory":
+            return {"entries": [
+                {"type": "file", "name": p.split("/")[-1], "path": p}
+                for p in file_map
+            ]}
+
+        if action == "read_file_content":
+            path = data.get("path", "")
+            return {"content": file_map.get(path, "")}
+
+        if action == "get_file_metadata":
+            path = data.get("path", "")
+            name = path.split("/")[-1]
+            ext = "." + name.rsplit(".", 1)[-1] if "." in name else ""
+            return {"name": name, "extension": ext, "size": 1024,
+                    "createdAt": None, "modifiedAt": None}
+
+        return {}
+
+    return _executor
+
+
+# ── Journey runner helper ─────────────────────────────────────────────────
+
+
+async def _run_journey(user_id: str, case: dict, executor) -> dict[str, Any]:
+    """Drive start + all user_messages for a case. Returns the final reply dict.
+
+    Mirrors ``device_ws._handle_journey_start/message``: sets the client
+    executor (so filesystem tools work) before each handler call.
+    """
+    session_id = str(uuid.uuid4())
+    try:
+        set_client_executor(executor)
+        reply = await handle_journey_start(user_id, {
+            "agent_type": "local",
+            "directory": case["directory"],
+            "data_types": case["data_types"],
+            "session_id": session_id,
+        })
+
+        for msg in case.get("user_messages", []):
+            if reply.get("done"):
+                break
+            set_client_executor(executor)
+            reply = await handle_journey_message(user_id, {
+                "session_id": reply["session_id"],
+                "message": msg,
+            })
+    finally:
+        clear_client_executor()
+        _sessions.pop(session_id, None)
+
+    return reply
+
+
+# ── Assertion helper ──────────────────────────────────────────────────────
+
+
+def _evaluate_case(case: dict, reply: dict) -> tuple[float, str]:
+    """Return (score, comment) for a journey case given the final reply dict."""
+    if case.get("expect_question"):
+        has_q = "?" in reply.get("message", "")
+        return (1.0 if has_q else 0.0), f"first_reply_has_question={has_q}"
+
+    if case.get("expect_done") and not reply.get("done"):
+        return 0.0, "expected done=True but journey did not complete"
+
+    agent_config_raw = reply.get("agent_config")
+
+    if case.get("expect_valid_config"):
+        if not agent_config_raw:
+            return 0.0, "agent_config is None"
+        try:
+            parsed = AgentConfig.model_validate_json(agent_config_raw)
+            valid = len(parsed.content_types) > 0
+            return (1.0 if valid else 0.0), f"content_types={len(parsed.content_types)}"
+        except Exception as exc:
+            return 0.0, f"parse error: {exc}"
+
+    if case.get("expect_content_type_id"):
+        expected_id = case["expect_content_type_id"]
+        if not agent_config_raw:
+            return 0.0, "agent_config is None"
+        try:
+            parsed = AgentConfig.model_validate_json(agent_config_raw)
+            ids = [ct.id for ct in parsed.content_types]
+            found = expected_id in ids
+            return (1.0 if found else 0.0), f"content_type_ids={ids}, expected={expected_id}"
+        except Exception as exc:
+            return 0.0, f"parse error: {exc}"
+
+    if case.get("expect_extraction_contains"):
+        keyword = case["expect_extraction_contains"].lower()
+        if not agent_config_raw:
+            return 0.0, "agent_config is None"
+        try:
+            parsed = AgentConfig.model_validate_json(agent_config_raw)
+            if not parsed.content_types:
+                return 0.0, "no content_types in config"
+            prompt = parsed.content_types[0].extraction_prompt.lower()
+            found = keyword in prompt
+            return (1.0 if found else 0.0), f"keyword='{keyword}' in extraction_prompt={found}"
+        except Exception as exc:
+            return 0.0, f"parse error: {exc}"
+
+    if case.get("expect_global_rules"):
+        if not agent_config_raw:
+            return 0.0, "agent_config is None"
+        try:
+            parsed = AgentConfig.model_validate_json(agent_config_raw)
+            has_rules = len(parsed.global_rules) > 0
+            return (1.0 if has_rules else 0.0), f"global_rules={parsed.global_rules}"
+        except Exception as exc:
+            return 0.0, f"parse error: {exc}"
+
+    return 1.0, "no specific assertion"
+
+
+# ── Unit tests ────────────────────────────────────────────────────────────
+
+
+def test_4_6a_extract_valid_json():
+    """_extract_agent_config: valid JSON between markers → returns serialised config."""
+    config = AgentConfig(
+        content_types=[],
+        global_rules=["No project = no entity"],
+        data_types=["tasks"],
+    )
+    text = f"Some preamble\n{_CONFIG_START}\n{config.model_dump_json()}\n{_CONFIG_END}\nTrailing"
+    result = _extract_agent_config(text)
+    assert result is not None
+    parsed = AgentConfig.model_validate_json(result)
+    assert parsed.global_rules == ["No project = no entity"]
+
+
+def test_4_6b_extract_invalid_json():
+    """_extract_agent_config: malformed JSON between markers → returns None."""
+    text = f"{_CONFIG_START}\n{{not: valid json\n{_CONFIG_END}"
+    assert _extract_agent_config(text) is None
+
+
+def test_4_6c_extract_markers_absent():
+    """_extract_agent_config: no markers at all → returns None."""
+    assert _extract_agent_config("No markers here at all") is None
+
+
+def test_4_6d_extract_only_start_marker():
+    """_extract_agent_config: START without END → returns None."""
+    assert _extract_agent_config(f"text {_CONFIG_START} no end marker") is None
+
+
+@pytest.mark.asyncio
+async def test_4_6e_session_not_found():
+    """4.6e Session not found → done=True, agent_config=None, informative message."""
+    reply = await handle_journey_message(_USER_ID, {
+        "session_id": "nonexistent-session-id",
+        "message": "Hello",
+    })
+    assert reply["done"] is True
+    assert reply["agent_config"] is None
+    assert "not found" in reply["message"].lower() or "expired" in reply["message"].lower()
+
+
+@pytest.mark.asyncio
+async def test_4_6f_nudge_uses_new_markers():
+    """4.6f Nudge injected after max turns uses AGENT_CONFIG markers, not PROMPT_TEMPLATE."""
+    session_id = str(uuid.uuid4())
+    captured_histories: list[list[dict]] = []
+
+    async def _mock_llm(system_prompt, history, tools, **kwargs) -> str:
+        captured_histories.append(list(history))
+        # Return plain text — no markers — to trigger the nudge path.
+        return "I still need more information from you."
+
+    from app.api.routes.agent_setup import JourneySession
+
+    fake_session = JourneySession(
+        session_id=session_id,
+        user_id=_USER_ID,
+        agent_type="local",
+        directory="/test",
+        data_types=["tasks"],
+        system_prompt="system",
+        langfuse_prompt=None,
+    )
+    # Fill history to the turn limit so the next message triggers the nudge.
+    for i in range(_MAX_TURNS):
+        fake_session.history.append({"role": "user", "content": f"msg {i}"})
+        fake_session.history.append({"role": "assistant", "content": "ok"})
+    _sessions[session_id] = fake_session
+
+    try:
+        with patch("app.api.routes.agent_setup._call_llm_with_tools", side_effect=_mock_llm):
+            await handle_journey_message(_USER_ID, {
+                "session_id": session_id,
+                "message": "one more message to trigger nudge",
+            })
+    finally:
+        _sessions.pop(session_id, None)
+
+    # Second LLM call receives the nudge appended to history.
+    assert len(captured_histories) >= 2, "Expected ≥ 2 LLM calls (main reply + nudge)"
+    nudge_history = captured_histories[1]
+    user_msgs = " ".join(t["content"] for t in nudge_history if t["role"] == "user")
+    assert _CONFIG_START in user_msgs, f"Nudge must reference {_CONFIG_START}"
+    assert _CONFIG_END in user_msgs, f"Nudge must reference {_CONFIG_END}"
+    assert "PROMPT_TEMPLATE" not in user_msgs, "Old PROMPT_TEMPLATE markers must not appear in nudge"
+
+
+# ── Eval tests (real LLM + Langfuse) ─────────────────────────────────────
+
+
+@pytest.mark.asyncio
+@pytest.mark.eval
+async def test_eval_journey(journey_case, pytestconfig):
+    """Parametrized eval test — one invocation per YAML case."""
+    case: dict = journey_case
+    fixtures_dir = _fixtures_dir(pytestconfig)
+    executor = _make_fs_executor(case.get("directory_files", []), fixtures_dir)
+
+    lf = get_langfuse()
+    obs_ctx = lf.start_as_current_observation(
+        name=f"eval-journey-{case['id']}-{case.get('score_name', 'unknown').replace('.', '-')}",
+        metadata={"step": "4", "case_id": case["id"]},
+    ) if lf else nullcontext()
+
+    with obs_ctx as obs:
+        reply = await _run_journey(_USER_ID, case, executor)
+        score, comment = _evaluate_case(case, reply)
+
+        if obs is not None:
+            obs.score(
+                name=case.get("score_name", f"journey.case_{case['id']}"),
+                value=score,
+                comment=comment,
+            )
+
+    if lf:
+        lf.flush()
+
+    assert score == 1.0, f"[{case['id']}] {case.get('description', '')} — {comment}"

tests/test_preprocessors.py

@@ -0,0 +1,98 @@
+"""Tests for the preprocessor system (Step 1 — Local Agent V2).
+
+Run:
+    pytest tests/test_preprocessors.py -v
+    pytest tests/test_preprocessors.py -v --preprocess-dir /path/to/folder
+
+The folder must contain cases.yaml + data/.
+"""
+
+from __future__ import annotations
+
+import re
+from pathlib import Path
+
+import pytest
+import yaml
+
+from app.core.preprocessors import detect_content_type, preprocess
+
+_DEFAULT_DIR = Path(__file__).parent / "fixtures" / "preprocessors"
+
+_GENERATORS = {
+    "binary_noise": "some\x00\x01\x02\x03\x04\x05content" * 20,
+}
+
+
+def _fixtures_dir(config) -> Path:
+    override = config.getoption("--preprocess-dir")
+    return Path(override) if override else _DEFAULT_DIR
+
+
+def _load_cases(config) -> list[dict]:
+    return yaml.safe_load((_fixtures_dir(config) / "cases.yaml").read_text(encoding="utf-8"))
+
+
+def _content(case: dict, data_dir: Path) -> str:
+    if "generate" in case:
+        return _GENERATORS[case["generate"]]
+    return (data_dir / case["file"]).read_text(encoding="utf-8")
+
+
+# ── parametrize at collection time via pytest hook ────────────────────
+
+def pytest_generate_tests(metafunc):
+    if "preprocess_case" not in metafunc.fixturenames:
+        return
+    cases = _load_cases(metafunc.config)
+    test_name = metafunc.function.__name__
+    if test_name == "test_detect":
+        subset = [c for c in cases if "detect" in c]
+    else:
+        subset = [c for c in cases if "process" in c]
+    metafunc.parametrize("preprocess_case", subset, ids=[c["id"] for c in subset])
+
+
+# ── detect ────────────────────────────────────────────────────────────
+
+def test_detect(preprocess_case, pytestconfig) -> None:
+    case = preprocess_case
+    data_dir = _fixtures_dir(pytestconfig) / "data"
+    raw = _content(case, data_dir)
+    filename = case.get("file", "")
+    ct = detect_content_type(filename, raw)
+    expected = case["detect"]
+    assert ct == expected, f"[{case['id']}] expected {expected!r}, got {ct!r}"
+
+
+# ── preprocess ────────────────────────────────────────────────────────
+
+def test_preprocess(preprocess_case, pytestconfig) -> None:
+    case = preprocess_case
+    data_dir = _fixtures_dir(pytestconfig) / "data"
+    raw = _content(case, data_dir)
+    result = preprocess(case["process"], raw)
+
+    if case.get("no_html"):
+        assert not re.search(r"<[^>]+>", result.clean_text), "clean_text contains HTML tags"
+
+    if "min_chars" in case:
+        assert len(result.clean_text) >= case["min_chars"], \
+            f"clean_text too short: {len(result.clean_text)} < {case['min_chars']}"
+
+    if "ratio_lt" in case:
+        ratio = len(result.clean_text) / len(raw)
+        assert ratio < case["ratio_lt"], f"compression ratio {ratio:.2f} >= {case['ratio_lt']}"
+
+    for key in case.get("has_meta", []):
+        assert result.metadata.get(key), f"metadata missing {key!r} (got {result.metadata})"
+
+    for item in ([case["contains"]] if isinstance(case.get("contains"), str) else case.get("contains", [])):
+        assert item in result.clean_text, f"clean_text missing {item!r}"
+
+    for item in ([case["excludes"]] if isinstance(case.get("excludes"), str) else case.get("excludes", [])):
+        assert item not in result.clean_text, f"clean_text contains forbidden {item!r}"
+
+    if "content_type" in case:
+        assert result.content_type == case["content_type"], \
+            f"expected content_type {case['content_type']!r}, got {result.content_type!r}"