菜单
Evaluate a 3rd-party Claude skill before importing — source trust, duplicate check, rule extraction, vixen-fit. Reject vague triggers and off-scope content. Use when the user shares a skill URL, asks "should we adopt this skill", or proposes importing from a marketplace.
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。
Bump the patch version in server/Cargo.toml or website/package.json after completing a major change. Use when the user asks to bump/release a version, or when CLAUDE.md's rule about version bumps after major changes applies.
Before a non-trivial change, list affected systems — DB schema, API contract, Telegram handlers, jobs, TS types, docs. Catch ripple effects upfront, not after CI fails. Use when planning a feature, migration, refactor, or any cross-module touch.
Self-review before push — read the diff as a reader, scan for vixen footguns (token leaks, PII, idempotency, captcha asset overwrites), run validation pipeline, no surprise files. Use before "git push", before opening a PR, or when the user says "ready to commit".
Write a git commit message for vixen-rs following the project's Conventional Commits convention. Use when the user asks to commit, create a commit, or write a commit message.
Reproduce → bisect → root cause → fix → verify. Don't speculate, don't "add more error handling" as a guess. Use when something is broken, when a test fails intermittently, when the user says "this isn't working" or "investigate why".
Write small, fast, reproducible Dockerfiles for the vixen-rs server (Rust) and website (bun) using multi-stage builds, cache mounts, distroless/alpine runtime images, and non-root users. Use when editing server/Dockerfile, website/Dockerfile, docker-compose.yml, or when the user asks to shrink an image or speed up a Docker build.
基于 SOC 职业分类
| name | find-external-skill |
| description | Evaluate a 3rd-party Claude skill before importing — source trust, duplicate check, rule extraction, vixen-fit. Reject vague triggers and off-scope content. Use when the user shares a skill URL, asks "should we adopt this skill", or proposes importing from a marketplace. |
Source: vixen meta + foxic discipline.
Imported skills must add a rule, not duplicate one. Skill quality over quantity. A repo with 50 vague skills is worse than 15 sharp ones.
anthropics/skills — Anthropic-curated. High trust.skills.sh, bencium-marketplace, accesslint-marketplace. Medium trust; verify per-skill.Before adopting, grep:
rg -l '<concept>' .claude/skills server/docs/rules website/docs/rules
If 80%+ of the rule already exists somewhere, skip the import. Or merge a single bullet into the existing skill instead of creating a duplicate.
description)The trigger must answer: "what user phrasing makes Claude pick this skill?"
Skills are rules + checklists, not essays. If a source skill is 2000 words of theory, rewrite down to:
Aim for ≤ 120 lines. Beyond that, you're writing a doc, not a skill.
i64, captcha assets immutable, redaction discipline (RedactedToken), idempotent moderation actions, transactional per-chat config writes. If the skill teaches a pattern that violates these, reject or rewrite.chat_config, moderation_actions, captcha_challenges).**Source:** line — required for attribution and re-checking..claude/skills/server/<name>/SKILL.md..claude/skills/website/<name>/SKILL.md..claude/skills/<name>/SKILL.md..claude/skills/infra/<name>/SKILL.md..claude/skills/README.md — name, trigger, area.**Source:** line.CLAUDE.md Critical Rules — re-read, don't assume.plan-before-code — adoption is itself a non-trivial change; plan first.import-skill (sister meta-skill, when added) — the mechanical "do the import" steps.