菜单
Runs checks from .ged/work/<work-id>/TESTS.md after task implementation, summarizes pass/fail, and prepares retry briefs. Triggers include "verify", "test", "check", "did it work", or after completing an implementation task.
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
Discover external skills during GedPi's skill-fit checkpoint. When the current bundled, project, and user skills don't cover a task's needs, search the open ecosystem via npx skills, evaluate candidates, and install through GedPi's project-skill mechanism. Triggers include "how do I do X", "find a skill for X", "is there a skill for X", "skill for", "need a skill", "search for skills", "skill-fit", "ecosystem"
Implements the next task from .ged/work/<work-id>/TASKS.md with narrow scope and concise handoff notes. Triggers include "do the next task", "implement", "execute", or when a task is ready for implementation.
Produces or refines implementation plans in .ged/work/<work-id>/SPEC.md, task breakdowns in .ged/work/<work-id>/TASKS.md, and verification criteria in .ged/work/<work-id>/TESTS.md. Triggers include "plan", "break down", "what tasks", "how to implement", or when starting a new feature.
Use before planning or implementing a non-trivial change when the request is not fully clear. Clarifies goals, desired behavior, constraints, edge cases, non-goals, tests, rollout, and success criteria by asking one concise question at a time with a recommended answer/default.
Grilling session that challenges your plan against the existing domain model, sharpens terminology, and updates documentation (CONTEXT.md, ADRs) inline as decisions crystallise. Use when user wants to stress-test a plan against their project's language and documented decisions. Triggers include "domain", "glossary", "terminology", "ubiquitous language", "CONTEXT.md", "ADR", "architecture decision", "decision record", "context map", "domain model".
Create a project-local skill when the skill-fit checkpoint finds no adequate existing skill. Use when find-skills returns nothing suitable, the task gap is reusable at the project level, and you need to capture domain knowledge, workflow patterns, or project-specific procedures as a SKILL.md. Triggers include "create a skill", "new skill", "write a skill", "make a skill for", "capture this", "turn this into a skill", "project skill", "custom skill"
| name | ged-verification |
| description | Runs checks from .ged/work/<work-id>/TESTS.md after task implementation, summarizes pass/fail, and prepares retry briefs. Triggers include "verify", "test", "check", "did it work", or after completing an implementation task. |
.ged/work/<work-id>/TESTS.mdFor non-trivial Ged tasks, planning should happen after classification, any needed grill-me clarification, and skill-fit resolution. With subagents enabled, ged-explorer performs read-only skill inventory/evaluation/search and reports findings; the main brain then installs project-scoped external skills or creates narrow project-local skills when warranted. In solo/no-subagent mode, the main brain performs the skill-fit checkpoint itself. Do not install global/user skills automatically.