菜单
Run the Goldfish Protocol — test a design doc's completeness by verifying a fresh session can reconstruct the plan from it alone. Use after /plan, before /implement, on any complex feature.
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
| name | egm |
| description | Run the Goldfish Protocol — test a design doc's completeness by verifying a fresh session can reconstruct the plan from it alone. Use after /plan, before /implement, on any complex feature. |
| model | opus |
| argument-hint | [path/to/design-doc.md or inline description] |
Implements the Elephant-Goldfish Model Goldfish check from Dave Rensin's "Elephants, Goldfish and the New Golden Age of Software Engineering."
A design doc is Goldfish-proof when a blank-slate session — with no conversation history, no context, only the written document — can reconstruct the plan accurately. This skill runs that check before implementation begins.
/plan produces a design doc or plan file/implement on any feature touching >3 filesAsk via dialog: When this skill needs a decision, preference, or clarification, call
AskUserQuestion(ToolSearch select:AskUserQuestionif the schema isn't loaded). Don't embed questions in prose. See~/.claude/CLAUDE.md§ "Asking the User Questions" for the full rule.
If an argument was passed, use it. Otherwise look for:
~/.claude-corp/plans/DESIGN.md or PLAN.md in the project root or docs/Spawn a sub-agent with only the design doc text as context. No codebase access. Ask it to answer:
Record what it gets right, what it gets wrong or vague, and what it had to guess.
Spawn a second independent sub-agent with only the design doc text. Ask it to act as a skeptical peer reviewer:
Synthesize both sub-agent results into a structured report:
EGM Check — [doc name or feature]
Comprehension Test:
✓ or ✗ What to build [pass / gaps found: ...]
✓ or ✗ Files affected [pass / gaps found: ...]
✓ or ✗ Acceptance criteria [pass / gaps found: ...]
✓ or ✗ Rejected alternatives [pass / gaps found: ...]
Critic Findings:
- [gap or ambiguity 1]
- [gap or ambiguity 2]
(empty = no issues found)
Session Recovery block present? yes / no
Verdict: GOLDFISH-PROOF ✓ — ready for /implement
or NEEDS REVISION ✗ — return to /plan to address gaps above
/implement/planAudit Ansible playbooks, roles, collections, and inventories for production readiness. Use when reviewing an Ansible repo, before merging IaC PRs, before promoting a role to a collection, or when the user mentions ansible-lint, idempotency, vault, or molecule.
Enter planning mode — interview the user, design a phased approach, and produce an implementation plan before writing code.
Execute an approved /plan unattended — phases, agents, commits — stopping only on hard safety failures (hook exit 2, sandbox/network deny, repeated test failures). Use after /plan when the user wants hands-off execution.
Run a full read-only audit of the current project — code review, security analysis, and documentation freshness check. Use to assess project health without making changes.
Start implementing a feature with user story validation, phased execution, and quality gates. Use after planning is complete.
Take a project from "it works" to "it's shippable." Runs audit, fixes findings, walks the Definition of Done checklist, and generates a release readiness score. Use before releasing, after feature-complete, or when quality feels prototype-y.