メニュー
Build or rebuild the ADR index and dependency graph in AgentDB
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Use when a conversation is getting long, drifting, juggling multiple tasks, re-pasting the same files, or showing stale/confused context — and at the start of any new task. Triggers: long thread, context bloat, topic switch, repeated file uploads, 'why is it forgetting', degraded answers, finished one task and starting another.
Use when a user wants to deconstruct a complex problem, business challenge, product decision, strategy question, career crossroads, or stuck situation by stripping assumptions, identifying first principles, rebuilding options from foundational truths, and choosing one high-leverage action.
Audit role-based access control and permission drift across product surfaces.
Add a new feature to an existing @fabrk/* package following monorepo conventions
Query AgentDB with semantic routing, hierarchical recall, causal graphs, and context synthesis
Use when the user wants to reduce AI coding spend without reducing shipping speed. Covers model routing, prompt caching, context discipline, loop profiling, and rollout plans for Cursor, Claude Code, Aider, and similar coding-agent stacks.
| name | adr-index |
| description | Build or rebuild the ADR index and dependency graph in AgentDB |
| argument-hint | |
| allowed-tools | mcp__claude-flow__agentdb_hierarchical-store mcp__claude-flow__agentdb_hierarchical-query mcp__claude-flow__agentdb_causal-edge mcp__claude-flow__agentdb_causal-query mcp__claude-flow__memory_store mcp__claude-flow__memory_search Bash Read Grep Glob |
Build or rebuild the full ADR index and dependency graph in AgentDB from the docs/adr/ directory.
After importing ADRs from another project, when the AgentDB graph is out of sync, or when bootstrapping ADR tracking on an existing codebase that already has ADR files.
Scan directory -- Glob for docs/adr/ADR-*.md to find all ADR files. If no files found, report that no ADRs exist yet.
Parse each ADR -- Read each file and extract:
ADR-042 from ADR-042-use-postgres.md)# ADR-NNN: <Title> heading**Status**: line**Date**: line**Tags**: line## Links section (supersedes, amended-by, related)Store in AgentDB -- For each ADR, call mcp__claude-flow__agentdb_hierarchical-store with:
adr/<adr-id>{ "id": "<id>", "title": "<title>", "status": "<status>", "date": "<date>", "tags": "<tags>", "file": "<filepath>" }Build causal edges -- For each ADR with links:
mcp__claude-flow__agentdb_causal-edge with from: ADR-XXX, to: <current>, relation: supersedesmcp__claude-flow__agentdb_causal-edge with from: <current>, to: ADR-YYY, relation: amendsmcp__claude-flow__agentdb_causal-edge with from: <current>, to: ADR-ZZZ, relation: relatedmcp__claude-flow__agentdb_causal-edge with from: <current>, to: ADR-WWW, relation: depends-onStore in memory -- For each ADR, call mcp__claude-flow__memory_store with:
adr-patterns<adr-id><title> — <first paragraph of Context section>
This enables semantic search across ADRs.Verify graph -- Call mcp__claude-flow__agentdb_causal-query to retrieve all edges and verify:
Report -- Output a summary:
## ADR Index Summary
Total ADRs: N
- Proposed: X
- Accepted: Y
- Deprecated: Z
- Superseded: W
Relationships: M edges
- Supersedes: A
- Amends: B
- Depends-on: C
- Related: D
Issues found: (list any dangling refs or status mismatches)