Twining MCP Server

Post an entry to the shared blackboard. Use this to share findings, needs, warnings, status updates, and other coordination messages with other agents. Does NOT accept entry_type 'decision' — use twining_decide instead.

Read blackboard entries with optional filters. Use this to check what other agents have posted, find relevant context, or review recent activity.

Semantic search across blackboard entries. Uses embeddings when available, falls back to keyword search. Returns entries ranked by relevance.

Get the most recent blackboard entries. Quick way to see latest activity without specifying filters.

Remove specific blackboard entries by ID. Use this to clean up false-positive warnings, resolved entries, or other noise. Returns which IDs were dismissed and which were not found.

Record a decision with full rationale, alternatives considered, and traceability. Creates a decision record and cross-posts to the blackboard.

Retrieve all decisions affecting a given scope or file. Shows the decision chain with rationale, confidence, and alternatives count. Essential for understanding "why was it done this way?"

Trace a decision's dependency chain upstream (what it depends on) and/or downstream (what depends on it). Uses BFS with cycle protection.

Flag a decision for reconsideration. Sets active decisions to provisional status and posts a warning to the blackboard with downstream impact analysis.

Override a decision with a reason. Sets the decision to overridden status, records who overrode it and why, and optionally creates a replacement decision automatically.

Promote one or more provisional decisions to active status. Use this to confirm provisional decisions that have been validated through implementation and testing.

Query decisions by commit hash. Returns all decisions that were linked to a given commit, enabling traceability from code changes back to decision rationale.

Search decisions across all scopes by keyword or semantic similarity. Returns ranked results without requiring a specific scope. Supports filtering by domain, status, and confidence level.

Link a git commit hash to an existing decision. Enables bidirectional traceability between decisions and commits.

Build tailored context for a specific task. Returns relevant decisions, warnings, needs, findings, and questions within a token budget. Call this before starting any task to get shared context from other agents.

Get a high-level summary of project or scope state. Returns counts of active decisions, open needs, warnings, and a recent activity narrative.

Report what changed since a given point in time. Returns new decisions, new entries, overridden decisions, and reconsidered decisions. Use this to catch up on changes since you last checked.

Run verification checks on a scope. Checks test coverage (tested_by relations), warnings acknowledgment, assembly-before-decision tracking, drift detection (P2 stub), and constraints (P2 stub). Auto-posts a finding with the summary.

Export full Twining state as a single markdown document. Includes blackboard entries, decisions with full rationale, and knowledge graph entities/relations. Use for handoff between context windows, documentation, or debugging.

List all registered agents with their capabilities and liveness status.

Register a new agent or update an existing one. Merges capabilities on re-registration. Use this to make subagents visible in the coordination dashboard.

Find agents matching required capabilities, ranked by capability overlap and liveness. Returns scored agent list for delegation decisions.

Post a delegation request to the blackboard as a 'need' entry with capability requirements. Returns suggested agents ranked by match quality.

Create a handoff record from one agent to another, capturing work results and auto-assembling context snapshot. Posts a status entry to the blackboard.

Acknowledge receipt of a handoff, recording which agent picked it up.

Overall health check of the Twining state. Shows blackboard entry count, decision counts, graph entity/relation counts, actionable warnings, and a human-readable summary.

Archive old blackboard entries. Moves entries older than a cutoff timestamp to an archive file, preserving decision entries. Optionally posts a summary finding.

Add or update a knowledge graph entity. Uses upsert semantics: if an entity with the same name and type exists, its properties are merged and updated. Returns the entity ID.

Add a relation between two knowledge graph entities. Source and target can be entity IDs or names. Returns an error for ambiguous name matches.

Traverse the knowledge graph from an entity, returning neighbors up to a given depth (max 3). Supports filtering by relation type. Useful for understanding how entities connect.

Search the knowledge graph for entities by name or property substring match. Case-insensitive. Returns matching entities with their properties.

Remove orphaned knowledge graph entities that have no relations. Use this to clean up stale or disconnected entities. Optionally filter by entity type to only prune certain kinds.