القائمة
Search cross-cycle memory: decisions, learnings, experience archives. Returns ranked matches.
يُحدَّد Prompt الذي يراجع المصدر أولًا بشكل افتراضي. يمكنك التبديل إلى أمر مباشر أو تنزيل نسخة محلية.
اقرأ SKILL.md وأي ملفات مرافقة يعرضها SkillsMP قبل أن تقرر التثبيت.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
يتجاوز الأمر المباشر Prompt المخصّص للمراجعة. افحص المصدر قبل تشغيله.
npx skills add https://github.com/hegemonart/get-design-done --skill gdd-recallيبقى الأمر في سطر واحد. مرّر أفقيًا لمراجعته كاملًا قبل النسخ.
تفضّل نسخة محلية؟ نزّل الملفات المتاحة حاليًا لدى SkillsMP.
Master design pipeline for Claude Code. 5-stage workflow: Brief → Explore → Plan → Design → Verify. Run 'brief' first in any new project to capture the design problem, then 'explore' to inventory the codebase and interview for context. Invoke without arguments for status and auto-routing.
Run the pipeline with optional agents skipped for speed. Skips: phase-researcher, design-assumptions-analyzer, design-integration-checker. Keeps: planner, executor, verifier, auditor. Activates for requests involving a lightweight design pass, a fast iteration, or a quick low-ceremony change.
Routes a /gdd command to fast|quick|full path + S|M|L|XL complexity_class and returns {path, complexity_class, model_tier_overrides, resolved_models, estimated_cost_usd, cache_hits}. A SKILL.md prompt the model executes to emit a routing-decision JSON from rule tables (no separate agent spawn). Optional/advisory - invoked only by the skills that opt into routing; the budget-enforcer hook tolerates its absence. Read by hooks/budget-enforcer.ts.
استنادا إلى تصنيف SOC المهني
عرض SKILL.md
| name | gdd-recall |
| description | Search cross-cycle memory: decisions, learnings, experience archives. Returns ranked matches. |
| argument-hint | <query> [--reindex] |
| tools | Read, Write, Bash |
| disable-model-invocation | true |
@reference/retrieval-contract.md
Searches .design/archive/, .design/learnings/LEARNINGS.md, .design/CYCLES.md, and STATE.md decision blocks for the given query and returns ranked matches. Uses FTS5 when better-sqlite3 is available, ripgrep next, Node fs scan as universal fallback.
Parse argument: extract <query> and optional --reindex flag.
--reindex: print usage and exit.--reindex only: rebuild the index (see step 3), then exit.Check backend: run
node -e "const {backendName}=require('./scripts/lib/design-search.cjs');console.log(backendName())"
Print a one-line notice: Search backend: fts5 | ripgrep | node-grep.
Reindex if requested: run
node -e "require('./scripts/lib/design-search.cjs').reindex(process.cwd())"
Print: "Index rebuilt."
Search: run
node -e "
const s=require('./scripts/lib/design-search.cjs');
const r=s.search(process.argv[1], process.cwd(), {limit:20});
console.log(JSON.stringify(r));
" -- "<query>"
Parse the JSON array of {file, line, text} objects.
Format results: print in this shape:
## Recall: "<query>" (N matches, backend: fts5)
### 1. .design/archive/cycle-2/EXPERIENCE.md:42
> Decided to skip dark-mode token layer in cycle 2 — cost–benefit did not clear threshold
### 2. .design/learnings/LEARNINGS.md:18
> L-03 · color · cited:4 · Color contrast failures cluster around interactive states, not surfaces
Group by file. Truncate text at 120 chars. Cap at 20 results.
Signal relevance: for every result whose text contains an L-NN learning ID, record a surfaced signal:
node -e "
const rc=require('./scripts/lib/relevance-counter.cjs');
rc.record('L-NN','surfaced','.design');
"
Empty result: if no matches, print:
No matches for "<query>" in cross-cycle memory.
Tip: run {{command_prefix}}recall --reindex if the archive has been updated since last index build.
.design/STATE.md.