Feat/goal scoped recommendations

[jv-singh]

Summary

Completes the Module 2 → Module 3 pipeline by connecting goal-scoped activity signals to the LangGraph recommendation engine. The AI now evaluates only the signals belonging to a specific goal and generates SMILE phase advancement recommendations.

• Store: Updates store.list_signals() in src/lpi/store.py to support optional goal_id filtering while preserving existing behavior for all current callers.
• LangGraph: Updates the recommendation prompt in src/lpi/langgraph_agent.py so the LLM analyzes activity signals for a single goal instead of the user's entire portfolio and recommends SMILE phase advancement when appropriate.
• API Route: Adds POST /api/v1/recommendations/{user_id}/by-goal/{goal_id} in src/lpi/routers/recommendations.py. Validates goal ownership, fetches goal-specific signals, runs LangGraph, and returns up to three prioritized recommendations.
• Security: Adds goal ownership validation to sync_github_events in src/lpi/routers/signals.py to prevent users from syncing GitHub activity to goals they do not own.
• Impact: Completes the Signals → Recommendations integration, unblocking Jahanvi (Frontend UI) and Daksh (QA & end-to-end verification).

Type

• Feature (new endpoint / model)
• Fix (bug, test failure)
• Refactor
• Docs / reports

Checklist

• make lint exits 0
• make test exits 0 (no new failures)
• Tests added or updated for changed behaviour
• Daily report updated in reports/
• No .env secrets committed
• PR targets staging (never main directly)

Tests passing

• tests/test_recommendations_goal_scoped.py (8 new tests)
• tests/test_recommendations.py (28 passed)
• tests/test_activity_signals.py (21 passed)
• tests/test_scoring.py (48 passed)
• tests/test_smoke.py (3 passed)

Context.docx