Skip to content

Latest commit

 

History

History
277 lines (246 loc) · 33.1 KB

File metadata and controls

277 lines (246 loc) · 33.1 KB

Skill Cheat Sheet

A scan-and-go "doing X → run this skill" map of every listed skill in this marketplace, generated from each skill's SKILL.md frontmatter by scripts/generate-cheatsheet.mjs. Do not hand-edit the generated block below. Edit the source frontmatter and regenerate.

Grouping axis: this page groups skills by sequence of use, the session-flow workflow stages you move through while working. The separate what-kind-of-plugin taxonomy axis is owned by docs/catalog-taxonomy.md.

0. Contract

Skill Plugin What it does
/bugs:write bugs Turn an informal bug description into a structured 5-field report, read-only
/planning:audit-answers planning Adversarially validate interview answers with fresh-context agents
/planning:brainstorm planning Diverge into codebase-grounded candidate approaches before scoping
/planning:interview planning Interview in frontier rounds until the task contract is locked
/planning:prd planning Lock product intent. Problem, users, success metrics. Before planning
/planning:questionnaire planning Turn a decision someone else must answer into an async questionnaire
/planning:wayfind planning Chart a too-big, foggy effort as a decision map worked one decision at a time

1. Explore

Skill Plugin What it does
/architecture:map-landscape architecture Chart a repository and the systems it references as a C4 system landscape and portfolio table
/discovery:blindspot discovery Surface your unknown-unknowns and sharpen the prompt before unfamiliar work
/discovery:explore discovery Explore code, history, tests, and config before changing anything
/discovery:trace-intent discovery Reconstruct why a thing was built this way, from evidence outside the code
/performance:target performance Rank optimization candidates by evidence quality, not suspicion

2. Research

Skill Plugin What it does
/context7:lookup context7 Look up current library docs, API references, and examples via Context7
/discovery:research discovery Multi-source external research with source tiers and a coverage ledger
/discovery:research-deep discovery Dispatch deep multi-topic research to the heaviest isolated tier
/dometrain:grounding dometrain Ground an approach in how a Dometrain course teaches it, with lesson links
/firecrawl:firecrawl firecrawl Scrape, search, crawl, or parse web pages when WebFetch is blocked

3. Plan

Skill Plugin What it does
/architecture:improve architecture Scan the codebase for shallow modules and friction, then design the chosen fix several ways
/domain-driven-design:curate-language domain-driven-design Maintain the domain glossary. Resolve terms, record rejected synonyms
/event-storming:methodology event-storming EventStorming facilitation reference across all three formats
/event-storming:simulation event-storming Multi-persona agentic EventStorming workshop on Miro
/naming:name-it-better naming Generate and evaluate name candidates from blind fresh-context lenses
/performance:goal performance Build a goal with realistic and ideal targets plus a computed floor
/planning:design planning Resolve types, contracts, and module boundaries before planning
/planning:design-handoff planning Gate a finished design and package it for planning
/planning:devils-advocate planning Stress-test a plan or the incumbent approach adversarially
/planning:draft-goal-condition planning Pick the right autonomy lever and craft a /goal completion condition
/planning:plan planning Produce a structured implementation plan with an approval gate
/prototype:explore-directions prototype Throwaway UI variations answering what should this look like
/prototype:pressure-test prototype Throwaway terminal app or shareable HTML demo pressure-testing logic or a data model
/work-items:decompose work-items Break a plan into vertical-slice work items with dependencies

4. Implement

Skill Plugin What it does
/debugging:debug debugging Diagnose broken behavior. Reproduce, hypothesise, instrument, fix with regression test
/implementation:implement implementation Execute approved plans with TDD, incremental validation, and green commits
/implementation:implement-dispatch implementation Orchestrate worker subagents to execute an approved plan
/instruction-placement:realign instruction-placement Apply accepted placement findings behind a per-item human gate
/overengineering:realign overengineering Execute accepted audit findings down the rollback ladder behind a per-item human gate
/source-control:commit source-control Commit with the resolved convention and surgical staging

5. Test

Skill Plugin What it does
/evals:design evals Interview for success criteria and scaffold an eval suite in the consumer repo
/evals:methodology evals Answer LLM-evaluation design questions from Anthropic's official guidance
/evals:plugin-eval evals Preflight, price, run, and read a plugin eval suite around the CLI
/evals:validate evals Static FAIL/WARN check of an eval suite's cases and graders, with no model call
/mutation-testing:audit mutation-testing Report surviving mutants on the diff, restoration verified or the run fails, survivors triaged
/mutation-testing:principles mutation-testing Answer mutation-testing questions from the primary literature
/playwright:playwright playwright Live E2E browser automation with disk-written artifacts
/tdd:principles tdd Answer test design questions from authoritative TDD sources
/testing:audit testing Detect tests that cannot fail. Report, gate, or persist
/testing:diagnose testing Root-cause failing tests, never retry blindly
/testing:plan testing Classify changes by required test type and coverage gaps
/testing:run-e2e testing Start the app, drive real flows, capture evidence
/testing:write testing Write and place tests with TDD cadence across ecosystems

6. Review

Skill Plugin What it does
/code-tidying:audit-comment-residue code-tidying Classify code comments for history narration and session-reference residue
/code-tidying:batch-simplify code-tidying Batch-run simplification across changed files, or a whole repository, by ecosystem
/code-tidying:dissolve-comments code-tidying Dissolve comments into expressive code via triage. Delete, refactor-then-delete, or keep
/mcp-tools:audit mcp-tools Audit MCP tool definitions against design quality criteria
/plugin-quality:audit plugin-quality Behavioral audit of a plugin component ending in a maintainer work item
/review:audit-enforceability review Propose the cheapest deterministic rung for each review finding
/review:code-review review Org CI code-review command for claude-review.yml
/review:fanout review Fan review out across every reviewer surface into one ranked report
/review:quality-gate review Single-lens review checkpoint routed to the matching reviewer
/review:security-review review Org CI security-review command for claude-security-review.yml
/skill-quality:check skill-quality Static QA gate for skill frontmatter, caps, and evals

7. Verify outcome

Skill Plugin What it does
/performance:snapshot performance Capture a snapshot only from a host proven measurable
/performance:verify performance Re-derive the result in fresh context and report it honestly
/toolchain:check toolchain Build, test, and lint changed files across detected ecosystems
/toolchain:lint toolchain Polyglot lint and format checks without a full build
/verification:confirm verification Prove the change achieved its intended outcome with evidence
/verification:measure verification Verify an improvement claim against a pre-change baseline

8. Retrospective

Skill Plugin What it does
/session-flow:retro session-flow Structured session retrospective with codified learnings
/session-flow:running-retro session-flow In-flight retro checkpoint appended to a running ledger

PR lifecycle (after step 7)

Skill Plugin What it does
/source-control:pull-request source-control Full PR lifecycle. Prep, create, monitor CI, address reviews, merge
/source-control:resolve-conflicts source-control Resolve merge and rebase conflicts by recovering both sides' intent

Anytime / cross-cutting

Skill Plugin What it does
/adhd:clarify adhd Reshape a dense, decision-heavy message into clear one-decision-at-a-time chunks, losing nothing
/adhd:shape adhd Set a standing action-first output posture. Lead with the next action, cut preamble
/ai-slop:audit ai-slop Detect and remove AI-writing tells from markdown prose
/architecture:record-decision architecture Record an architecture decision in the repository's existing ADR convention
/claude-config:audit claude-config Audit settings, hooks, permissions, and MCP config for drift against current official docs
/claude-config:audit-automation-gaps claude-config Audit the repo's automation landscape for hook, MCP, skill, and subagent gaps worth adding
/claude-config:audit-instructions claude-config Find instructions current models no longer need across CLAUDE.md, rules, and skill bodies
/claude-config:audit-pass claude-config Run one coordinated, resumable audit pass over a repo with a single human gate
/claude-config:audit-permission-grants claude-config Audit permission grants for portability and auto-mode durability
/claude-config:audit-permission-state claude-config Report the permission rules actually in effect and what auto mode drops
/claude-config:audit-prompting-postures claude-config Find posture guidance the prompting guide says a component needs but does not carry
/claude-config:draft-auto-mode-rules claude-config Interview and draft a paste-ready autoMode block, never writing settings
/claude-config:unhobble claude-config Strip instructions to a bare baseline, log real stumbles, re-add only what evidence earns
/claude-memory:audit claude-memory Audit CLAUDE.md, rules, and auto-memory against the official-docs checklist
/claude-memory:stateless claude-memory Inspect, disable, or purge Claude Code's per-repo auto memory
/claude-ops:changelog claude-ops Ingest a Claude Code release changelog and integrate its changes into the repo
/claude-ops:known-issues claude-ops Look up and track known Claude product issues, health, and workarounds
/code-metrics:audit-complexity code-metrics Per-function complexity beside a cited reference, no verdict
/code-metrics:audit-coverage code-metrics Coverage and CRAP read from build artifacts, no verdict
/code-metrics:audit-duplication code-metrics Clone groups minus the replication the repo declares, no verdict
/code-metrics:audit-size code-metrics Lines per file beside a cited reference, no verdict
/code-metrics:audit-type-debt code-metrics Typed-code percentage per file and per lane, with no standard behind it
/code-metrics:principles code-metrics What each code measure can and cannot tell you
/code-tidying:audit-dead-code code-tidying Whole-repo dead-code hunt across four labelled lanes with adjudicated candidates
/code-tidying:tidy code-tidying Proactively hunt one lane for safe structural tidyings and ship a structure-only PR
/codebase-health:audit codebase-health Audit for drift between docs, config, code, and architecture via verified findings
/computer-use:diagnose computer-use Resolve computer-use capture, input, and screenshot symptoms to a cause
/context-budget:audit context-budget Measure the startup context payload per item and ledger every lever's real delta
/coupling:reduce coupling Scan for change-transmitting coupling, apply safe reductions in a budgeted batch, route the rest
/discipline:do-your-research discipline Re-anchor research discipline, then audit and correct the current work
/discipline:do-your-research-deep discipline Verify every session claim against primary sources in a heavy fan-out
/discipline:follow-our-standards discipline Re-anchor to org engineering standards and audit the work in flight
/discipline:hold-my-hand discipline Standing posture. Phase the work, present one phase at a time, wait for the advancing reply
/discipline:mind-your-maxims discipline Re-anchor cooperative communication and audit recent responses for clarity
/discipline:pick-for-the-problem discipline Re-derive a tool or approach choice from the problem, not habit
/discipline:point-dont-copy discipline Audit for copied content and correct by pointing at the living source
/discipline:reason-dont-recite discipline Challenge decisions coasting on precedent, re-derive from first principles
/discipline:recheck-against-upstream discipline Audit the surface in flight against current official upstream docs
/discipline:recheck-against-upstream-deep discipline Fan out doc-by-doc upstream conformance checks across a whole subsystem
/discipline:reuse-or-replace discipline Reuse the established way or openly replace it, never a silent second way
/discipline:script-the-deterministic-work discipline Script counting, diffing, and transforms instead of eyeballing them
/discipline:scrutinize-dont-coast discipline Re-examine your own recent output through a fresh-context pass
/discipline:sweep-all discipline Batch every discipline corrector into one audited re-anchor pass
/discipline:tighten-your-output discipline Tighten prose and code. Fewer words, no semantic loss
/discipline:use-your-skills discipline Map the task to available skills and invoke them instead of reinventing
/discipline:wait-what discipline Re-pitch the message that did not land. Missing context added, plain register, project vocabulary
/disk-hygiene:clean disk-hygiene Audit a directory tree for stale leftovers and remove validated paths
/docs-hygiene:audit-derivability docs-hygiene Judge whether a doc earns its existence or should become a pointer
/docs-hygiene:audit-encapsulation docs-hygiene Find external citations reaching into a skill's private surfaces
/docs-hygiene:audit-file-names docs-hygiene Inventory a doc tree's file names and plan the renames with their references
/docs-hygiene:audit-noise docs-hygiene Classify markdown for citations, ghost refs, meta-commentary, plan/conversational/tracker residue
/docs-hygiene:audit-progressive-disclosure docs-hygiene Grade instruction files for split opportunities and hub/spoke disclosure defects
/docs-hygiene:compress docs-hygiene Tighten markdown by dropping flavor while preserving every directive
/docs-hygiene:extract-ssot docs-hygiene Deduplicate repeated prose into one named source of truth
/docs-hygiene:realign-file-names docs-hygiene Apply a file-name rename plan, one acceptance per file
/docs-hygiene:rename-references docs-hygiene Sweep stale references after renames, including forms grep misses
/docs-hygiene:write-for-agents docs-hygiene Authoring-time doctrine for agent-consumed markdown
/docs-hygiene:write-for-humans docs-hygiene Authoring-time doctrine for human-read documentation
/education:eli5 education Visual HTML explainer assuming zero prior knowledge, one idea per diagram
/education:explain education Explain any concept or the last response in genuinely plain words
/education:quiz-me education Generate a post-change report with a quiz verifying you absorbed the work
/education:teach education Multi-session learning coach for general topics or repo-grounded concepts
/fleet:reach fleet Run a Claude Code agent turn on another fleet machine over SSH
/github:advise github Design and set up GitHub settings and admin areas grounded in live gh state
/github:audit github Read-only audit of GitHub org and repo settings, drift, and cost signals
/improvement:find improvement Rank evidence-cited improvement candidates across dimensions; execution goes to the pipeline
/instruction-placement:audit instruction-placement Find instruction content on the wrong surface and propose validated destinations
/instruction-placement:check instruction-placement Gate that every path-scoped rule glob resolves and the rules index is current
/instruction-placement:delta instruction-placement Report only what moved since the last placement audit
/overengineering:audit overengineering Audit the enforcement surface for mechanisms no longer earning their carry cost
/overengineering:delta overengineering Re-run the enforcement-surface audit and report only what moved since the last run
/overengineering:justify overengineering Make one artifact you point at justify its own existence, on evidence
/playbooks:boris playbooks Boris Cherny's Claude Code workflow tips across 115 sections
/playbooks:fable-5 playbooks Fable 5's operating doctrine loaded as standing session instructions
/playbooks:skill-authoring playbooks Anthropic's internal skill-authoring playbook and patterns
/playgrounds:use playgrounds Route playground requests to the first-party plugin, or install it, with recipes and guidance
/provenance:audit provenance Find prose copied from external sources and convert it into pointers
/repo-hygiene:clean repo-hygiene Clean caches, build artifacts, stale branches, and stashes per repo
/session-flow:show-options session-flow Lay out the skills that fit this moment as a ranked, nothing-hidden menu
/session-flow:workflow session-flow Navigate the staged dev workflow and suggest the next stage
/visualization:visualize visualization Pick the best visual form for what is in the conversation and render it
/wizard:generate wizard Author a hardened interactive bash wizard for human-only setup, credential, and cutover steps
/work-items:scan-todos work-items Sweep source comments for TODO and FIXME markers, resolve or file each
/work-items:ship work-items Route a spec container's macro journey. Status, execution shape, next step
/work-items:track work-items Backlog CRUD through the bound tracker. Add, list, close, stats
/work-items:triage work-items Evaluate raw intake through the verified-to-eligible state machine
/work-items:work work-items Auto-select one work item and execute it end-to-end
/writing:be-concise writing Reshape prose for a scanning reader. Bottom line first, about half the words

Session lifecycle

Skill Plugin What it does
/session-flow:clean-stop session-flow Make everything durable before the machine goes away
/session-flow:continue-in-background session-flow Delegate the task to a fresh background agent now
/session-flow:find-handoff session-flow Recover a lost handoff or resume prompt after /clear
/session-flow:handoff session-flow Write a mid-session save-point for clear-and-resume
/session-flow:keep-going session-flow Recover after an interruption and continue where work stood
/session-flow:orchestrate session-flow Arm the session with proactive-orchestration imperatives
/session-flow:orient session-flow Read-only situation report from durable and off-thread state
/session-flow:reanchor session-flow Verify working assumptions are still true before building on them
/session-flow:reconcile session-flow Retire finished off-thread work and square the task ledger
/source-control:worktree source-control Create, inspect, and clean git worktrees for parallel sessions

Operator cadence

Skill Plugin Cadence What it does
/bugs:scan bugs daily Proactively hunt resting code for unobserved bugs, verify adversarially, report read-only
/claude-ops:audit-install-state claude-ops weekly Audit a Claude Code install directory. What is there, what the product manages, what is stale
/claude-ops:audit-native-overlap claude-ops weekly Map native Claude Code surfaces against this repo's components and record human-gated verdicts
/claude-ops:audit-performance claude-ops continuous Capture slowness evidence while slow. Version, sweep health, tree walk, sessions, fleet, fan-out
/claude-ops:audit-skill-visibility claude-ops weekly Which skills the model can actually see, which are starved, and which are unobservable
/claude-ops:inventory claude-ops weekly Enumerate every command, skill, agent, and plugin component this machine can invoke
/claude-ops:lanes claude-ops daily Start, restart, stop, and check loop lanes as named background sessions
/claude-ops:morning-brief claude-ops daily Print the operator's read-only morning view. Queues, merge-ready PRs, parked decisions
/claude-ops:observability claude-ops weekly Report on locally captured telemetry. Token burn, cost, hook latency, per-session activity
/claude-ops:plugins claude-ops weekly Bring the machine's plugin fleet current. Refresh, update, install per policy
/repo-fleet-hygiene:apply repo-fleet-hygiene weekly Execute a fleet action plan behind one confirmation gate
/repo-fleet-hygiene:audit repo-fleet-hygiene weekly Discover a repository fleet and coordinate read-only evidence handoffs
/source-control:babysit-loop source-control continuous Run one repo's PR queue as a standing merge lane
/source-control:babysit-prs source-control continuous Tiered fleet pass advancing your open PRs
/work-items:attend-queue work-items daily Drive escalated and untriaged items to resolution in one view
/work-items:work-loop work-items continuous Drain the backlog as a self-paced autonomous loop