代码评审
自动化代码审查与分析
sd-profile
platypeeps
Use when the user wants to create, inspect, correct, review, import, export, or forget a consent-driven personal operating profile with traceable assertions.
sd-docs-bustest
platypeeps
Use when documentation must survive a cold read — checking that a newcomer with no prior context can execute a README, runbook, setup guide, or handoff doc exactly as written, when docs are created or changed, or when asked to bus-test docs.
sd-adr-review
platypeeps
Use when a PR or diff touches docs/adr/, DECISIONS.md, or *.adr.md files, when an ADR moves between proposed, accepted, rejected, or superseded status, or when the user asks to review an architecture decision record; checks MADR-style completeness, RFC-2119 driver force, honest consequences, forward links, lifecycle validity, and premise freshness, and reports P1/P2/P3 findings with one verdict line.
sd-rust-review
platypeeps
Use when a diff, branch, or pull request touches Rust — .rs files, Cargo.toml, lint configuration — to run the pack's Rust-specific probe checklist as a local lens whose findings feed the review of record.
sd-coherence-audit
platypeeps
Use when a knowledge corpus — a note vault, agent-instruction files, or a docs tree — must be audited against itself for contradictions, vagueness, bandaid guidance, and redundancy, returning a read-only findings ledger with both sides quoted.
sd-watchlist
platypeeps
Use when the user wants a read-only review of configured channels, feeds, authors, searches, playlists, podcasts, or collections that reports only material new items since an explicit checkpoint.
sd-typed-holes
platypeeps
Use when starting a Rust feature, module, or rewrite skeleton-first — design the types and signatures, land a compiling skeleton whose bodies are todo!() as its own commit, then fill the holes in a separate later pass; never mix the two.
sd-propose-skills
platypeeps
Use when the user wants the current session reviewed for recurring friction, repeated steps, and hard-won gotchas, and high-bar skill proposals drafted into the report for the user to decide.
sd-evaluate
platypeeps
Use when the user wants one defined subject assessed against an explicit rubric with criterion-level evidence, uncertainty, sensitivity, deficiencies, and prioritized improvements.
sd-meeting-follow-through
platypeeps
Use when the user wants a source-traceable post-meeting package that reconciles intended and actual outcomes, decisions, commitments, unresolved items, and consent-gated follow-through.
algo-sensei
moido1092
Your personal DSA & LeetCode mentor. Use for problem explanations, progressive hints, code reviews, mock interviews, pattern recognition, complexity analysis, and custom problem generation. Automatically adapts to your learning style and request type.
money-change
Munyon-Canyon
Checklist for any backend change that moves or records money, including ledgers, shares, swaps, transfers and balances. Use before writing or reviewing code that touches USDC, token amounts, share math, the relayer or Privy signing.
hmasd-research-engineering
CartmanFatass
Implement, delegate, check, review, execute and publish HMASD direction results under docs/project/OPERATING_CONSTITUTION.md - L0 scope, Implementer handoff, core versus disposable code, engineering standards, independent review, native launch and DM publication of its own RESEARCH entry. Not for status or mechanical collection.
analyze
Mike-Byrne-AI
Ground or re-ground the harness on this repo. On first run (no saved fingerprint) it establishes the baseline — what this codebase is, its stack, architecture, and test posture; thereafter it detects drift. Use when grounding the harness on a new repo, when project structure has shifted significantly, when returning to the project after time away, when docs/CONVENTIONS.md or docs/SHARP_EDGES.md feels stale, or when the user asks to analyze, fingerprint, ground, or check for changes. Delegates the heavy lifting to the repo-analyst and docs-maintainer agents.
w-check-diagnostics
reardan
Diagnose W compile errors and warnings with the compiler's structured check mode instead of full compiles. Use when a .w file fails to compile, when the edit hook reports diagnostics, or before running any tests on modified W code.
to-spec
CartmanFatass
"Turn the current conversation into a spec and publish it to the project issue tracker: no interview, just synthesis of what you've already discussed."
code-review-expert
mj9733246-cloud
"Expert code review of current git changes with a senior engineer lens. Detects SOLID violations, security risks, and proposes actionable improvements."
hmasd-loop-dispatch
CartmanFatass
Coordinate user-requested Codex Root assignments and cross-direction controls under the HMASD constitution. Independent DMs publish their own results and handle concurrent changes; messages between Codex App tasks require an explicit user request. Status questions and workflow edits do not resume research.
go-concurrency
Munyon-Canyon
Chooses between Pool, Stage, FanOut and errgroup for concurrent Go backend work, and applies the eight concurrency rules. Use when backend code starts goroutines, uses channels, or fans work out over providers, wallets or cabals.
ask-matt
CartmanFatass
Ask which skill or flow fits your situation. A router over the skills in this repo.
arch-check
Mike-Byrne-AI
Review changes for layer boundary violations, import direction, circular dependencies, and hook-exit-code errors. Use when refactoring across modules, when adding a new module, when a hook script imports espalier (forbidden), or when the user asks about architectural consistency. Delegates to the architecture-analyst agent.
hmasd-loop-dispatch
CartmanFatass
Coordinate user-requested Codex Root assignments and cross-direction controls under the HMASD constitution. Independent DMs publish their own results and handle concurrent changes; messages between Codex App tasks require an explicit user request. Status questions and workflow edits do not resume research.
hmasd-scientific-tools
CartmanFatass
Develop cumulative HMASD research understanding - update judgments from evidence, use simple-model prototypes and select a discriminating next action with independent scientific review under constitution section 5; add Pro when it offers distinct value. Declare fit cost and read confirmation by its fixed rule in NOTES.md.
design
Mike-Byrne-AI
Harness configuration review. Use when reviewing harness configuration, auditing the agent roster, evaluating the command set, checking hook coverage, or assessing docs freshness — also when the user asks to review setup, audit harness, check for stale agents, or when significant project changes warrant a configuration review. Delegates to the harness-config-advisor agent.