Extends timeline() with Jira integration since sessions moved to Jira-only in
CF-836 (2026-02-08) and the session-mcp tables are empty by design.
Adds to services/jira.ts:
- getIssueWithHistory(key): fetches issue with expand=changelog + comments
- searchIssueKeys(jql): JQL search returning minimal issue keys
- ADF → plain text extractor for comment bodies
Timeline now yields Jira events: issue_created, field_change:{status,assignee,
resolution,priority,labels}, comment. Events are time-filtered client-side
against the since/until window. For Jira-key subjects, also searches for
linked session-tracking issues and merges their events.
Tested against CF-2872 (audit task) and CF-2885 (this ticket) — shows full
lifecycle from creation through transitions to resolution.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
New `timeline(subject, since, until)` tool stitches events across session-mcp
sources (sessions, notes, commits, plans, task-commit links) into a single
time-ordered stream for LLM consumption.
Phase 1 of CF-2885: read-time stitching only, no business-system changes.
Per CF-2830 heise article thesis (Event Sourcing + MCP for LLMs).
Subject accepts Jira key, session id, or project key. Optional sources filter,
since/until window (supports relative -7d shorthand), limit.
Returns markdown-formatted timeline grouped by date with per-source icons.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Hybrid (vector + keyword + rerank) search over indexed session transcripts.
Enables context recovery from past sessions without re-reading JSONL files.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Extract structured metadata (topics, decisions, blockers, tools_used,
projects, issue_keys) from session summaries using Haiku at session end.
Metadata stored in JSONB column with GIN index for filtered retrieval.
session_semantic_search now accepts optional metadata filters.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add rerank() function calling LiteLLM /v1/rerank endpoint (Cohere-compatible).
Plugged into all 3 search functions (sessions, session-docs, archives) after
RRF merge. Disabled by default via RERANK_ENABLED env var. Graceful fallback
to RRF-only ranking on API failure.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Auto memory (MEMORY.md + topic files) is the sole memory system.
The session-mcp memory_* tools (PostgreSQL + pgvector) had zero entries
after months of use — removing dead code to simplify the server.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Wrap CallToolRequest handler with withSentryTransaction for
per-tool tracing. Remove broken $(vault) DSN from .env.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add PostgreSQL full-text search alongside pgvector for exact matches
on Jira keys, error messages, file paths. Merge results with
Reciprocal Rank Fusion. Default mode: hybrid, with graceful
degradation to keyword-only when embeddings unavailable.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
SHA-256 hash check before embedding API call eliminates ~60-80% of
redundant embedding requests. Consolidates dual INSERT paths to single
INSERT with nullable embedding column.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add gateway-first pattern: when AGILITON_API_KEY is set, route all
external API calls through the gateway with X-API-Key auth. Falls back
to direct API access when gateway is unavailable.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Tables migrated to Jira tracking. PostgreSQL tables will be dropped separately.
Archived in s3://macbookdev/db-archive/agiliton-db-2026-02-08.sql.gz
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The project parameter was passed to createSessionIssue() but ignored -
all sessions were created in the CF Jira project regardless of the
actual session project (ST, LLB, GB, etc.).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
session_sequences table fell behind when sessions were inserted with
explicit session_number (e.g., retro imports), causing duplicate key
violations on next auto-assigned number. Function now syncs forward
by checking MAX(session_number) before assigning.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Sessions now auto-create CF Jira issues on start and post full session
output as comments on end, transitioning the issue to Done.
- Add src/services/jira.ts with createSessionIssue, addComment, transitionToDone
- Update session_start to create CF Jira issue and store key in sessions table
- Update session_end to post session output and close Jira issue
- Add migration 031 to archive local task tables (moved to Jira Cloud)
- Update .env.example with Jira Cloud env vars
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
When session_id is not provided, falls back to getSessionId() which
reads from CLAUDE_SESSION_ID env or ~/.cache/session-memory/current_session.
Fixes NOT NULL constraint violation on session_notes.session_id.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Adds DB column, TypeScript types, MCP tool schemas, and CRUD handlers
for planning_mode_required (NULL=auto-detect, true=always, false=never).
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Remove invalid 'pending' status from task_list and task_update enums
- Default to excluding completed tasks when no status filter provided
- Previously, task_list(status=open) missed in_progress/blocked/testing tasks
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Phase 3 cleanup: Transition complete to transcript-based recovery.
- Remove fallback to recoverNotesFromTempFile()
- All recovery now uses JSONL transcripts only
- Log warning if transcript not found for recovery
- Simplify recovery logic (no more dual-method path)
This completes the daemon elimination:
- LaunchAgent plist: Deleted
- session-notes-sync script: Deleted
- Legacy fallback code: Removed
- Result: Simpler architecture, no daemon overhead
Migration complete. System now uses only transcript-based recovery.
Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
- Migration 025: Add 'abandoned' status to sessions CHECK constraint (fixes blocking issue)
- Migration 026: Add recovery metadata columns (recovered_from, recovered_at) to track note recovery source
- Update sessionRecoverOrphaned to recover notes from temp files when marking sessions abandoned
- Update notes-parser to track recovery source and timestamp for analytics
These changes complete Priority 3, 5 and part of Priority 1 for CF-572 Session Notes Loss fix.
Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
- Add PROJECT_KEY_REGEX for valid format (2-5 uppercase letters)
- Add validateProjectKey() and isValidProjectKey() functions
- Update getProjectKey() to validate input and generated keys
- Reject invalid formats with clear error messages
Invalid formats now rejected:
- Single letters (A, C, U)
- Numbers (1, 20, 123)
- Full names (ClaudeFramework, Circles)
- Mixed case (Circles)
- Too long (>5 chars)
Also fixes Sentry SDK v8 API changes (httpIntegration, postgresIntegration).
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Create sentry.ts with MCP-aware initialization and PII scrubbing
- Replace basic inline Sentry initialization with initSentry()
- Update .env.vault-mapping for sentry.backend-node.dsn secret
- Includes PostgreSQL integration and transaction tracing
Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
Fixes task project rename command by adding ON UPDATE CASCADE to all foreign key constraints that reference projects(key).
Updated 10 FK constraints: task_sequences, epic_sequences, versions, epics, tasks, project_locks, project_archives, sessions, session_sequences, project_documentation.
Migration drops old constraints and adds new ones with both ON DELETE CASCADE and ON UPDATE CASCADE (except sessions which uses ON DELETE SET NULL).
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Added 'done' status to task lifecycle for verification period before completion.
**Changes:**
- Added 'done' to task_status enum (pending, open, in_progress, testing, blocked, done, completed)
- Added auto_close_at timestamp column for scheduling auto-close
- Created index for efficient cron job queries
**Workflow:**
1. Task marked 'done' → auto_close_at = NOW() + 7 days
2. Cron job runs daily to check tasks past auto_close_at
3. If no related issues → auto-close to 'completed'
4. If related issues → keep as 'done' (prevents premature closure)
**Lifecycle:**
open → in_progress → done (7-day verification) → completed (auto-closed)
task:CF-454
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Enhanced task_add to search both open AND completed tasks when checking for
duplicates. Previously only checked non-completed tasks.
**Changes:**
- Removed `status != 'completed'` filter from similarity search
- Increased search limit from 3 to 5 tasks
- Separated display: open tasks vs previously completed
- Show 150-char context snippet from completed task descriptions
- Suggest using "task show <id>" to see full solution
**Benefits:**
- Prevents recreating already-solved tasks
- Surfaces solutions from previous attempts
- Reduces circular work on recurring issues
- Maintains context from closed tasks
**Example output:**
```
⚠️ Similar tasks found:
**Open/In Progress:**
- CF-442: Next.js cache issue (92% match, open)
**Previously Completed:**
- CF-378: Pre-test validation (85% match)
Context: "Created comprehensive pre-test quality validation workflow..."
💡 Use "task show <id>" to see full solution before recreating work
```
Fixes: CF-450
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Extends project_documentation with 'workflow' type for:
- Development workflows
- Quality validation processes
- Testing procedures
Used for CF-378 pre-test validation workflow.
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Extends project_documentation with 'configuration' type to store:
- Public/internal domains
- Component locations (VMs, services)
- Dependencies and infrastructure setup
**Changes:**
- Added 'configuration' to doc_type enum in tools/index.ts
**Use case:**
Displayed at session start via session-start script (CF-382)
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
New link types added:
- depends_on: Task depends on another task to complete first
- needs: Task needs another task (weaker dependency)
- implements: Task implements a feature/requirement
- fixes: Task fixes an issue identified in another task
- causes: Task causes/introduces issue in another task
- subtask_of: Task is a subtask of another (parent-child)
Preserves existing types: blocks, relates_to, duplicates
Changes:
- Migration 021 updates database constraint
- Updated TypeScript types and MCP tool schema
- Applied to production database
Implements CF-385
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
- DB_HOST: 10.0.1.1 → infra.agiliton.internal
Improves maintainability and enables dynamic DNS resolution.
Relates to CF-382
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Changed hardcoded port 5432 to read from POSTGRES_PORT environment
variable to support pgbouncer connection pooling (port 6432).
Connection pooling reduces overhead for frequent queries and better
handles concurrent connections.
.env now configured with POSTGRES_PORT=6432 (pgbouncer).
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
- Check if session exists before inserting commit link
- Auto-create minimal session record if missing (status=active, no project)
- Prevents FK constraint violation when cache/DB drift occurs
- Return message indicates if session was auto-created for monitoring
- Handles edge cases: concurrent calls, missing committed_at, NULL project
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Resolves MCP error when trying to set status="pending" on tasks.
Changes:
- Migration 019: Add 'pending' to tasks.status CHECK constraint
- Update task_update tool schema to accept 'pending' status
- Update task_list tool schema to filter by 'pending' status
Status meanings:
- pending: Created, awaiting approval or dependencies
- open: Approved and ready to start
- in_progress: Currently being worked on
- testing: Implementation complete, being tested
- blocked: Stuck waiting for something
- completed: Done
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Removed hardcoded internal IP address from debug log to comply with
CF-303 IP exposure validation. Changed to generic message.
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Added validation to check if session_id exists in sessions table before
inserting memories. Falls back to NULL if session not found, preventing
foreign key constraint violations while preserving memory content.
Logs warning when invalid session_id provided for debugging.
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Added task_move_project tool to move tasks between projects while
preserving all relationships and history:
Changes:
- db.ts: Added getClient() for transaction support
- crud.ts: Implemented taskMoveProject() with full transaction
- index.ts: Wired up tool handler and import
- tools/index.ts: Registered tool definition
Preserves:
- Task metadata (title, description, type, status, priority)
- Checklist items
- Linked commits
- Delegations
- Activity history
- Task links (relates_to, blocks, duplicates)
- Epic and version assignments
- Embeddings
Process:
1. Validates source task and target project exist
2. Generates new ID in target project
3. Transfers all related data in transaction
4. Marks old task as completed with reference
5. Creates duplicate link between tasks
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
