aa-ma-search
Search across active and completed AA-MA task files for keywords. Returns ranked results from reference.md, context-log.md, and tasks.md files. Use when the user asks to search past AA-MA tasks.
- License 许可
- Apache-2.0
- 最近更新
- 2026-11-11
- Stars
- ★ 6
- Forks
- 1
01 前言 Front Matter
- name
- aa-ma-search
- description
- Search across active and completed AA-MA task files for keywords. Returns ranked results from reference.md, context-log.md, and tasks.md files. Use when the user asks to search past AA-MA tasks.
03 SKILL.md 正文
/aa-ma-search — Cross-Task Search
Search across all AA-MA task files to find past decisions, facts, and task state.
Usage
/aa-ma-search <query> [--active|--completed|--all]
<query>— keyword or phrase to search for--active— search only.claude/dev/active/(default)--completed— search only.claude/dev/completed/--all— search both active and completed tasks
Instructions for AI
Step 1: Parse Arguments
Extract the search query and scope from the user's input.
QUERY = [everything after /aa-ma-search except flags]
SCOPE = --active (default) | --completed | --all
If no query provided, use AskUserQuestion to prompt: "What would you like to search for across AA-MA tasks?"
Step 2: Determine Search Directories
# Build list of directories to search
SEARCH_DIRS=""
case "$SCOPE" in
--active|"")
# Check project-level first, then HOME
if [ -d ".claude/dev/active" ]; then
SEARCH_DIRS=".claude/dev/active"
elif [ -d "$HOME/.claude/dev/active" ]; then
SEARCH_DIRS="$HOME/.claude/dev/active"
fi
;;
--completed)
if [ -d ".claude/dev/completed" ]; then
SEARCH_DIRS=".claude/dev/completed"
elif [ -d "$HOME/.claude/dev/completed" ]; then
SEARCH_DIRS="$HOME/.claude/dev/completed"
fi
;;
--all)
# Combine both, checking project-level first
for subdir in active completed; do
if [ -d ".claude/dev/$subdir" ]; then
SEARCH_DIRS="$SEARCH_DIRS .claude/dev/$subdir"
elif [ -d "$HOME/.claude/dev/$subdir" ]; then
SEARCH_DIRS="$SEARCH_DIRS $HOME/.claude/dev/$subdir"
fi
done
;;
esac
If no directories found, report: "No AA-MA task directories found. Nothing to search."
Step 3: Execute Search
Search across reference.md, context-log.md, and tasks.md files within each task directory. Use Grep tool for efficient searching.
Target files (in each task directory):
*-reference.md— immutable facts*-context-log.md— decisions and trade-offs*-tasks.md— execution state and acceptance criteria
Do NOT search:
*-provenance.log— machine telemetry, not human-readable knowledge*-plan.md— duplicates content in tasks.md and reference.md
Step 4: Rank and Present Results
Present results grouped by task, with match count for basic relevance ranking:
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
AA-MA Search: "[query]"
Scope: [active|completed|all]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
? [task-name-1] (N matches)
? reference.md:
- [matching line with context]
? context-log.md:
- [matching line with context]
? [task-name-2] (N matches)
? tasks.md:
- [matching line with context]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Total: N matches across M tasks
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Step 5: Handle Empty Results
If no matches found:
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
AA-MA Search: "[query]"
Scope: [scope]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
No matches found.
Suggestions:
- Try broader search terms
- Use --all to include completed tasks
- For semantic search (concept matching), see future roadmap
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Limitations
This is a keyword search command. It finds exact word matches, not conceptual similarity. For example, searching "authentication" will NOT find entries about "OAuth2 PKCE flow" unless the word "authentication" appears in the text.
Semantic/vector search is planned as a future enhancement when task volume warrants it (see specification Section XIII).
04 同仓库更多 Skills
deploy-helper
You can use this to help with deploy stuff, deploying, deployments, releases, shipping, pushing to prod, going live, rollouts.
writing-for-agents
Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.
verify-plan
Slash entry for plan verification: runs plan-verification on an existing AA-MA plan, writes [task]-verification.md, and with --iterate loops until the plan converges. Use when the user types /verify-plan or wants the verification report saved.
verify-impl
Post-impl adversarial review symmetric to /verify-plan. Dispatches up to 5 parallel audit agents (code-reviewer, security-auditor, tdd-sequence-auditor, context7-evidence-auditor, future-proofing-auditor) based on the milestone's plan-declared Audit-Profile. CRITICAL findings surface via AskUserQuestion accept/dispute/defer panel before §7.3 user authorization. Introduced in v0.8.0 per ADR-0005. Invoked by Phase 6.8 of /execute-aa-ma-milestone.
understand-codebase
Onboard to a new, inherited, or shared codebase. Read it, understand it, map it, and learn its conventions, versioning, tests, tech stack, rules (AGENTS.md / CLAUDE.md / .cursorrules / etc.), build/run/CI, integrations, security posture, and repo health — then produce an honest pros/cons verdict, a "contribute safely" playbook, and an "add a feature" playbook, written to ONBOARDING.md at the repo root plus a .claude/onboarding/ set of deep-dives — and optionally author an AGENTS.md if one is missing, or review and propose improvements to an existing one. Tiered (Quick / Standard / Deep). Reuses codemem (index + diagrams), gsd-map-codebase, system-mapping, code-intelligence, impact-analysis rather than re-implementing them; Deep tier runs a TeamCreate agent-team. Keywords: new codebase, shared codebase, inherited code, onboard, understand this repo, how do I contribute, how do I add a feature, ramp up, get oriented, ONBOARDING.md, AGENTS.md, codebase walkthrough, joining a project.
token-compression
Output token compression for AA-MA workflows. Reduces response tokens by 40-75% while preserving full technical accuracy. Three intensity levels mapped to execution modes: lite (HITL), full (general), ultra (AFK). Auto-activates during AA-MA execution based on task Mode field. Manual activation via /compress lite|full|ultra.