菜单
SkillsMP 已收集 ClayGendron/vfs 中的 10 个 Skill。打开任一 Skill 可查看来源和详情。
已展示 10 / 10 个已收集 Skill。
Run the storage conformance suite against real database engines (Postgres, MySQL, MSSQL, Oracle) in local Docker containers, end to end — start Docker Desktop, build the containers up, run the engine legs, tear everything down, quit Docker Desktop. Use when…
原文语言：英语
Review the tests of a commit set (or a named code area) — not the code — asking whether they actually pin the contract the code declares. Use when the user asks to "review the tests", "check test coverage" of a change, asks "would a test catch this", wants to…
原文语言：英语
Test the full library end-to-end — the fast inner loop on one Python, the full CI-parity matrix (scripts/ci.sh, Python 3.11–3.14 with lint/format/types/coverage) to verify "this is green" before commit and push, and the real-engine database legs. Use when the…
原文语言：英语
Run the full multi-agent code review of a commit set — one commit or a contiguous range — or an explicitly named code area; five parallel review agents (ownership_review, contract_review, scale_review, test_review, adversarial_review), an adversarial verifier…
原文语言：英语
Teach the user a part of this codebase (a file, module, or concept) as a paced, multi-response learning journey using Diátaxis as the framing. Use when the user asks to "learn", "understand", or "teach me" something, or invokes /teach. This file also…
原文语言：英语
Attack a commit set (or a named code area) single-handedly by writing and executing throwaway scripts against it, with an executed-repros-only evidence standard. Use when the user asks to "adversarially review", "attack", "pressure test", or "try to break" a…
原文语言：英语
Adversarially verify an implementation against its declared contracts — docstrings, specs, ADRs, type signatures, commit messages, and error-classification tables — catching both broken promises and behavior no contract mentions. Use when the user asks to…
原文语言：英语
Review the code a commit set touched (or a named code area) for logic that has accumulated in the wrong place, decomposing it into concerns and assigning each to its rightful owner, with duplication as the primary evidence. Produces an ownership map and…
原文语言：英语
Review a commit set (or a named code area) for correctness and boundedness at production scale — 10,000+-entry batches, the least generous SQL engine, real concurrency. Use when the user asks to "scale review", "check this at scale", "will this survive a 10k…
原文语言：英语
Adversarially verify one code-review finding by trying to refute it — reproduce the claimed failure, check declared intent, and check reachability — before it is allowed to reach the user. Use when a review agent hands you a single finding to verify, when the…
原文语言：英语