القائمة
Thin Claude adapter for spec-harness implement-task on one .spec feature TASK.
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/icecreamovo147/smart-recruit --skill run-agent-taskيبقى الأمر في سطر واحد. مرّر أفقيًا لمراجعته كاملًا قبل النسخ.
تفضّل نسخة محلية؟ نزّل الملفات المتاحة حاليًا لدى SkillsMP.
Review current Git branch, working tree, staged/unstaged/untracked changes, or one specific commit on the current branch like a senior code review lead, then generate paired Chinese review reports in both self-contained HTML and Markdown. When actionable findings exist, also emit a code-review-sdd.md repair plan for the code-review-sdd skill. Use when the user asks to inspect current branch changes, review a specific commit, compare against base branches or commits, produce Code Review Leader reports, create HTML/MD review reports similar to a provided sample, summarize architecture/risk/test impacts, or generate actionable code-review findings.
Execute repair tasks from a Code Review Leader `code-review-sdd.md` plan, serially fixing CR-* findings with revalidation, scoped edits, tests, and evidence. Use when the user asks to apply a code-review-sdd plan, fix Code Review Leader findings, run code-review-sdd, continue/resume review repairs, or remediate issues from `.code-review-sdd/code-review-sdd.md`.
Perform an evidence-based, read-only audit of the Smart Recruit `.knowledge/` layer against the current repository, identify stale or conflicting knowledge and coverage gaps, and generate exactly three synchronized Chinese deliverables: a Markdown audit report, a self-contained HTML audit report, and an agent-ready Markdown remediation plan. Use when Codex is asked to audit or review the project knowledge base, check whether `.knowledge` matches current code, find outdated/stale knowledge, assess knowledge drift or route coverage, or produce a knowledge repair plan.
استنادا إلى تصنيف SOC المهني
عرض SKILL.md
| name | run-agent-task |
| description | Thin Claude adapter for spec-harness implement-task on one .spec feature TASK. |
Use this adapter only to execute one canonical .spec/<feature-name> TASK.
Required input:
Feature: <feature-name>Task: <TASK-ID>Canonical mapping:
spec-harness
Mode: implement-task
Feature: <feature-name>
Task: <TASK-ID>
Required behavior:
AGENTS.md..agents/skills/spec-harness/SKILL.md.prompts/implement-task.md..spec/<feature-name>/reports/.Do not create provider-private task state, branch policy, completion rules, or review verdicts. Do not commit, push, merge, or update local Claude permissions.