一键导入
search
Search across HQ and indexed repos (qmd-powered semantic + full-text). Falls back to Grep if qmd is unavailable.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Search across HQ and indexed repos (qmd-powered semantic + full-text). Falls back to Grep if qmd is unavailable.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Generate ASCII block-art banner images for social posts and OG images. Dark midnight aesthetic with faded {company} text on near-black background. Two sizes — 1080x1080 (social) and 1200x630 (OG). Use for product announcements, project launches, concept visuals, and repo OG images.
Explore approaches and tradeoffs before committing to a PRD. Research HQ context, compare options, surface unknowns, generate brainstorm.md with recommendation.
Post-ship documentation sync — updates README, CLAUDE.md, architecture docs, and INDEX files to match what actually shipped. Use after merging a PR, completing a project, or deploying a release. Triggers on "update docs", "sync documentation", "docs are stale", "document what shipped".
Execute a single PRD story through coordinated worker phases (Ralph pattern). Each worker handles its domain, passes context to the next, with back-pressure (tests/lint/typecheck) keeping code on rails.
Prepare for a new session to continue this work. Captures session learnings, syncs domain knowledge and insights, commits dirty repos, writes a thread file and handoff.json, updates INDEX files, and refreshes the search index. Ensures continuity across sessions.
Capture and classify learnings, route to structured policy files (rules) or insight files (educational knowledge). Deduplicates via qmd (Grep fallback), rebuilds policy digest after policy changes. Callable manually or from /execute-task, /run-project, /handoff, /checkpoint. Use --hard flag for hard-enforcement rules.
| name | search |
| description | Search across HQ and indexed repos (qmd-powered semantic + full-text). Falls back to Grep if qmd is unavailable. |
| allowed-tools | Read, Grep, Bash(qmd:*), Bash(grep:*), Bash(ls:*) |
Semantic + full-text search across HQ and indexed codebases using qmd. Falls back to Grep if qmd is unavailable.
Input: The user's search query, with optional flags.
Extract from the user's input:
query — search text (everything except flags)--mode — search (BM25), vsearch (semantic), query (hybrid). Default: search-n — result count (default: 10)-c — collection name (e.g. hq, {product}). Default: auto-detect or all collections--full — show full content of top resultIf -c was NOT explicitly provided, infer the active company from context:
companies/{name}/ or repos/private/ matching a company repo per companies/manifest.yaml → use that company's collection/run {worker} is active and worker has company: field → use that company's collectionAvailable collections: hq (all HQ), {product} ({PRODUCT} codebase), {company}, {company}, personal, {company}, {company}, {company}, {company}, {company}, {company}, {company}, {company}, {company}, {company}, {company}, {company}
When auto-detected, display: (auto: {company}) in results header.
which qmd 2>/dev/null && qmd --version 2>/dev/null
If qmd is not available, skip to Fallback section.
Run the matching qmd command. Add -c $COLLECTION if a collection was specified or auto-detected:
Default (BM25 full-text):
qmd search "$QUERY" -n $N --json [-c $COLLECTION]
Semantic (conceptual match):
qmd vsearch "$QUERY" -n $N --json [-c $COLLECTION]
Hybrid (BM25 + vector + re-rank — best quality, slower):
qmd query "$QUERY" -n $N --json [-c $COLLECTION]
Parse JSON output. Display:
Search: "{query}" (mode: {mode}, collection: {collection or "all"})
Results:
1. [0.92] hq: knowledge/public/Ralph/02-core-concepts.md
"Ralph methodology emphasizes small loops with human checkpoints..."
2. [0.84] {product}: libs/core/src/auth/middleware.ts
"export function authMiddleware..."
3. [0.71] hq: workers/public/dev-team/architect/skills/design-review.md
"Architecture review following Ralph back-pressure patterns..."
{n} results. Use --full to show top result content.
qmd://{collection}/ prefix)If --full flag, after listing results, read the top result file with the Read tool.
If qmd is unavailable or errors:
Use the Grep tool to search file contents:
knowledge/, companies/, workers/, .claude/commands/, workspace/Display: "qmd unavailable, falling back to Grep"
If Grep is also unavailable, run:
grep -rl "$QUERY" ~/HQ/knowledge/ \
~/HQ/companies/ \
~/HQ/workers/ \
~/HQ/.claude/commands/ \
~/HQ/workspace/ 2>/dev/null | head -20
search ralph # BM25 keyword search (default, all collections)
search "how do workers execute" --mode vsearch # Semantic across all
search auth middleware -c {product} # Search {PRODUCT} codebase only
search "webhook handler" -c {product} --mode vsearch # Semantic search in {PRODUCT}
search {company} brand --mode query # Hybrid with re-ranking
search stripe -n 20 # More results
search authentication --full # Show top match content
search "brand guidelines" -c {company} # Search {company} knowledge only
search "recovery metrics" -c {company} # Search {Product} knowledge only
# If cwd is companies/{company}/:
search "case study" # Auto-detects → -c {company}
search mode is fastest — use for exact keywords--mode vsearch for conceptual/semantic queries--mode query for highest quality (slower, uses LLM re-ranking)-c to scope to a collection: hq, {product}, {company}, {company}, personal, {company}, {company}, {company}, {company}, {company}, + 7 more (run qmd status for full list)-c, auto-detects company from context; falls back to all collections/search-reindex after adding new content