Skip to content

Latest commit

 

History

History
411 lines (369 loc) · 97.9 KB

skill-cheat-sheet.md

File metadata and controls

411 lines (369 loc) · 97.9 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. The first table, By kind of task, is the Task kinds table of /session-flow:workflow's stage definitions: the stages each kind of task enters, in order, each linked to the table of skills that run it, with the command to type for that stage, what those commands run for you (each claim citing the skill line that makes the call), and the stages it skips. The command per stage and its claims are STAGE_COMMANDS in scripts/cheatsheet-config.mjs; the arguments each command takes are in its stage table. In the stage tables, each row's Actions are the action words of the skill's argument-hint (the first argument's choices), Arguments is the whole hint, Upstream links the upstream attribution table that lists the skill, and Next lists the skills its ## Next section names, in order; a cell is empty when the skill has none. Do not hand-edit the generated block below. Edit the source SKILL.md or the Task kinds table 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.

By kind of task

From the stage definitions: A conditional stage runs only when its trigger holds; otherwise it is skipped and marked SKIPPED, with the reason, on its checklist box. Stages 2-4 are detours: the interview, or a later stage, calls one when a question needs it, and the work returns to the stage that called it.

Kind of task Commands to type, in order What they run for you Skips Why
Small fix Implement: /implementation:implement (when the breakage is not yet diagnosed, first /debugging:debug) → Test: /testing:write (when a test fails, /testing:diagnose) → Review: /review:quality-gate → Verify outcome: /verification:confirm → Ship: /source-control:pull-request /verification:confirm: runs the build, test and lint checks through /toolchain:check and /toolchain:lint (plugins/verification/skills/confirm/SKILL.md:36); /verification:confirm: finds the touched features with /app-control:drive affected (plugins/verification/skills/confirm/SKILL.md:166); /verification:confirm: drives the touched features with /testing:run-e2e (plugins/verification/skills/confirm/SKILL.md:168); /source-control:pull-request: ready runs the code and security reviews in fresh-context subagents and the verify gate before marking the PR ready (plugins/source-control/skills/pull-request/SKILL.md:73) Interview, Explore, Research, PRD, Design, Plan, Decompose The quick-change on-ramp: the diff is quick to review and cheap to retry, so reading the code it needs is part of implement; a diff that turns out not to be quick to review goes back to the interview
Bug Interview: /planning:interview → Explore (conditional: a question needs the code read): /discovery:explore → Research (conditional: a question rests on an external claim): /discovery:research → Design: /planning:design → Plan: /planning:plan → Decompose (conditional: split the plan into tickets): /work-items:decompose → Implement: /implementation:implement (when the breakage is not yet diagnosed, first /debugging:debug) → Test: /testing:write (when a test fails, /testing:diagnose) → Review: /review:quality-gate → Verify outcome: /verification:confirm → Ship: /source-control:pull-request /verification:confirm: runs the build, test and lint checks through /toolchain:check and /toolchain:lint (plugins/verification/skills/confirm/SKILL.md:36); /verification:confirm: finds the touched features with /app-control:drive affected (plugins/verification/skills/confirm/SKILL.md:166); /verification:confirm: drives the touched features with /testing:run-e2e (plugins/verification/skills/confirm/SKILL.md:168); /source-control:pull-request: ready runs the code and security reviews in fresh-context subagents and the verify gate before marking the PR ready (plugins/source-control/skills/pull-request/SKILL.md:73) PRD The full path less the PRD, whose skip conditions name bug fixes; design records a one-line early exit for a single-file fix, and undiagnosed breakage is diagnosed before implement
Feature Interview: /planning:interview → Explore (conditional: a question needs the code read): /discovery:explore → Research (conditional: a question rests on an external claim): /discovery:research → PRD (conditional: lock product intent): /planning:prd → Design: /planning:design → Plan: /planning:plan → Decompose (conditional: split the plan into tickets): /work-items:decompose → Implement: /implementation:implement (when the breakage is not yet diagnosed, first /debugging:debug) → Test: /testing:write (when a test fails, /testing:diagnose) → Review: /review:quality-gate → Verify outcome: /verification:confirm → Ship: /source-control:pull-request /verification:confirm: runs the build, test and lint checks through /toolchain:check and /toolchain:lint (plugins/verification/skills/confirm/SKILL.md:36); /verification:confirm: finds the touched features with /app-control:drive affected (plugins/verification/skills/confirm/SKILL.md:166); /verification:confirm: drives the touched features with /testing:run-e2e (plugins/verification/skills/confirm/SKILL.md:168); /source-control:pull-request: ready runs the code and security reviews in fresh-context subagents and the verify gate before marking the PR ready (plugins/source-control/skills/pull-request/SKILL.md:73) None The full path
Refactor Interview: /planning:interview → Explore (conditional: a question needs the code read): /discovery:explore → Research (conditional: a question rests on an external claim): /discovery:research → Design: /planning:design → Plan: /planning:plan → Decompose (conditional: split the plan into tickets): /work-items:decompose → Implement: /implementation:implement (when the breakage is not yet diagnosed, first /debugging:debug) → Test: /testing:write (when a test fails, /testing:diagnose) → Review: /review:quality-gate → Verify outcome: /verification:confirm → Ship: /source-control:pull-request /verification:confirm: runs the build, test and lint checks through /toolchain:check and /toolchain:lint (plugins/verification/skills/confirm/SKILL.md:36); /verification:confirm: finds the touched features with /app-control:drive affected (plugins/verification/skills/confirm/SKILL.md:166); /verification:confirm: drives the touched features with /testing:run-e2e (plugins/verification/skills/confirm/SKILL.md:168); /source-control:pull-request: ready runs the code and security reviews in fresh-context subagents and the verify gate before marking the PR ready (plugins/source-control/skills/pull-request/SKILL.md:73) PRD Engineering-internal work, which the PRD's skip conditions name
UI change Interview: /planning:interview → Explore (conditional: a question needs the code read): /discovery:explore → Research (conditional: a question rests on an external claim): /discovery:research → PRD (conditional: lock product intent): /planning:prd → Design: /planning:design → Plan: /planning:plan → Decompose (conditional: split the plan into tickets): /work-items:decompose → Implement: /implementation:implement (when the breakage is not yet diagnosed, first /debugging:debug) → Test: /testing:write (when a test fails, /testing:diagnose) → Review: /review:quality-gate → Verify outcome: /verification:confirm → Ship: /source-control:pull-request /verification:confirm: runs the build, test and lint checks through /toolchain:check and /toolchain:lint (plugins/verification/skills/confirm/SKILL.md:36); /verification:confirm: finds the touched features with /app-control:drive affected (plugins/verification/skills/confirm/SKILL.md:166); /verification:confirm: drives the touched features with /testing:run-e2e (plugins/verification/skills/confirm/SKILL.md:168); /source-control:pull-request: ready runs the code and security reviews in fresh-context subagents and the verify gate before marking the PR ready (plugins/source-control/skills/pull-request/SKILL.md:73) None The full path; design starts from the flow and the screens, and review adds an interface review
Release Verify outcome: /verification:confirm → Ship: /source-control:pull-request /verification:confirm: runs the build, test and lint checks through /toolchain:check and /toolchain:lint (plugins/verification/skills/confirm/SKILL.md:36); /verification:confirm: finds the touched features with /app-control:drive affected (plugins/verification/skills/confirm/SKILL.md:166); /verification:confirm: drives the touched features with /testing:run-e2e (plugins/verification/skills/confirm/SKILL.md:168); /source-control:pull-request: ready runs the code and security reviews in fresh-context subagents and the verify gate before marking the PR ready (plugins/source-control/skills/pull-request/SKILL.md:73) Interview, Explore, Research, PRD, Design, Plan, Decompose, Implement, Test, Review The changes being released went through their own stages; what is left is proving the release candidate and shipping it

1. Interview

Skill Plugin What it does Actions Arguments Upstream Next
/bugs:write bugs Turn an informal bug description into a structured 5-field report, read-only [--unattended] [--file] [--quick|--full] [--no-survey] <bug description> /work-items:track, /debugging:debug, /implementation:implement
/planning:audit-answers planning Adversarially validate interview answers with fresh-context agents [topic] /planning:design
/planning:interview planning Interview in frontier rounds until the task contract is locked me, lock, scope, restate, share, document, product [me|lock|scope|restate|share|document|product] [topic] Pocock to-questionnaire, Pocock batch-grill-me, Pocock domain-modeling /planning:design, /planning:audit-answers, /planning:wayfind, /session-flow:workflow
/planning:wayfind planning Chart a too-big, foggy effort as a decision map worked one decision at a time chart, work [chart|work|<item link>] [topic|map] [item] Pocock wayfinder /planning:wayfind, /planning:interview, /planning:plan, /work-items:decompose, /planning:prd

2. Explore

Skill Plugin What it does Actions Arguments Upstream Next
/architecture:map-components architecture Chart the modules inside one deployable as a C4 component view [container] [--group-by directory|namespace|layer] [--layers <list>] [--dialect likec4|c4-plantuml] /architecture:improve, /planning:design, /architecture:map-landscape, /architecture:record-decision
/architecture:map-containers architecture Chart deployables, the stores they bind, and contained modules [system] [--dialect likec4|c4-plantuml] [--out <dir>] /architecture:map-components, /architecture:map-flow, /architecture:map-landscape, /planning:design, /architecture:record-decision
/architecture:map-context architecture Chart one software system's C4 system context from tracked configuration [system] [--actors <file>] [--focal <name>] [--dialect likec4|c4-plantuml] [--out <dir>] /architecture:map-landscape, /architecture:record-decision, /architecture:improve, /planning:design
/architecture:map-data architecture Draw an ERD from a committed schema, offline [--scope module|all|<module>] [--include-columns] [--dialect mermaid|dbml] [--out <dir>] /architecture:record-decision, /planning:design, /architecture:map-landscape
/architecture:map-dependencies architecture Cite project and package edges from build manifests [path] [--include-external] [--external-only] [--cycles-only] [--out <dir>] /architecture:map-components
/architecture:map-deployment architecture Chart IaC deployment topology per environment, with a diff [environment] [--diff <env-a> <env-b>] [--dialect likec4|c4-plantuml] [--out <dir>] /architecture:record-decision, /planning:design, /architecture:map-landscape
/architecture:map-events architecture Chart publishers, consumers, and orphan messages [--unrouted-only] [--out <dir>] /architecture:map-flow, /architecture:map-containers, /architecture:record-decision, /planning:design
/architecture:map-flow architecture Trace one C# entry point into a sequence diagram with a citation on every hop <entry point> [--depth N] [--out <dir>] /architecture:map-events, /architecture:record-decision, /planning:design
/architecture:map-landscape architecture Chart a repository and the systems it references as a C4 system landscape and portfolio table [--repos <path>[,<path>...]] [--root <dir>] [--out <dir>] [--check] [--remote[=all]] /architecture:map-components, /architecture:improve, /discovery:explore, /architecture:record-decision
/debugging:analyze-profile debugging Read a profile, trace or heap snapshot down to a file:line cause [<artifact path>] pstack runtime-forensics /debugging:debug, /performance:goal
/discovery:blindspot discovery Surface your unknown-unknowns and sharpen the prompt before unfamiliar work [area-or-domain] /discovery:explore, /planning:interview, /session-flow:workflow
/discovery:explore discovery Explore code, history, tests, and config before changing anything [--output change-prep|explain] [scope] pstack how /discovery:research, /planning:interview, /planning:design, /session-flow:workflow
/discovery:trace-intent discovery Reconstruct why a thing was built this way, from evidence outside the code <target> pstack why /planning:design, /planning:plan
/performance:go-faster performance Whole-process speed sweep with evidenced, ranked findings and adopt-now speedups [--unattended] /performance:goal, /overengineering:audit, /harness-ops:audit-performance, /performance:target
/performance:target performance Rank optimization candidates by evidence quality, not suspicion [<path|component|'session'|'telemetry'>] pstack perf-issue /performance:goal, /performance:go-faster

3. Research

Skill Plugin What it does Actions Arguments Upstream Next
/context7:lookup context7 Look up current library docs, API references, and examples via Context7 lookup, update [lookup <library> <query>|update]
/discovery:read-docs discovery Read an upstream docs page through the shared lookup and cache <url-or-slug> [question] /discovery:research, /planning:plan
/discovery:research discovery Multi-source external research with source tiers and a coverage ledger [breadth=low|medium] [deep] [--unattended] [topic] /planning:interview, /planning:design, /session-flow:workflow
/dometrain:grounding dometrain Ground an approach in how a Dometrain course teaches it, with lesson links [grounding query]
/firecrawl:firecrawl firecrawl Scrape, search, crawl, or parse web pages when WebFetch is blocked scrape, search, crawl, map, parse, interact, agent, monitor, developer, research, doctor <scrape|search|crawl|map|parse|interact|agent|monitor|developer|research|doctor|…> [args]

4. PRD

Skill Plugin What it does Actions Arguments Upstream Next
/planning:brainstorm planning Diverge into codebase-grounded candidate approaches before scoping <rough-problem> /planning:interview, /planning:prd, /planning:design, /session-flow:workflow
/planning:prd planning Lock product intent. Problem, users, success metrics. Before planning [tier] [task description|#<map>] pstack principle-experience-first, Pocock to-spec /planning:interview, /planning:design, /work-items:track, /session-flow:workflow

5. Design

Skill Plugin What it does Actions Arguments Upstream Next
/planning:design planning Resolve types, contracts, and module boundaries before planning library, module, data, integration, system [library|module|data|integration|system] [status|thread <name>|discuss|terminology|handoff] pstack typescript-best-practices, pstack architect, pstack principle-boundary-discipline, pstack principle-exhaust-the-design-space, pstack principle-experience-first /planning:plan, /planning:design-handoff, /user-interface:design
/planning:design-handoff planning Gate a finished design and package it for planning /planning:plan, /planning:design
/user-experience:structure user-experience Shape user flows, information architecture and content structure, evidence-labeled [flow, navigation or app context] /user-interface:design
/user-interface:design user-interface Design interfaces from the project's system and installed tools; terminal guidance built in [what to design] /prototype:explore-directions, /user-interface:write-css, /user-interface:write-animation, /user-interface:audit

6. Plan

Skill Plugin What it does Actions Arguments Upstream Next
/architecture:improve architecture Scan for shallow modules, boundary crossings or unwritten conventions, then hand on the fix deepening, boundaries, conventions [deepening|boundaries|conventions] [--medium <value>] pstack arena, pstack architect, pstack principle-boundary-discipline, pstack principle-laziness-protocol, pstack principle-minimize-reader-load, Pocock domain-modeling, Pocock improve-codebase-architecture, Pocock codebase-design /planning:plan, /coupling:reduce, /corrections:codify, /architecture:enforce-boundaries
/architecture:improve-layout architecture Check whether files that change together live together, then propose a target layout [--path <dir> ...] [--<setting> <value> ...] /planning:plan, /architecture:improve
/domain-driven-design:ubiquitous-language domain-driven-design Find, read, seed, maintain and check the domain glossary locate, lookup, seed, curate, audit [locate|lookup|seed|curate|audit] [term, ambiguity, or resolved vocabulary] /planning:plan
/event-storming:methodology event-storming EventStorming facilitation reference across all three formats [--big-picture|--process|--design-level|--patterns|--glossary|--notation|--remote] /event-storming:simulation, /planning:design
/event-storming:simulation event-storming Multi-persona agentic EventStorming workshop on Miro <--simulate|--process-model|--design-level|--evaluate|--retrospective…> [domain|board] /domain-driven-design:ubiquitous-language, /planning:design
/naming:name-it-better naming Generate and evaluate name candidates from blind fresh-context lenses [tournament] pstack arena
/performance:goal performance Build a goal with realistic and ideal targets plus a computed floor [<target>] pstack hillclimb /performance:snapshot
/planning:devils-advocate planning Stress-test a plan or the incumbent approach adversarially deep, shallow [deep|shallow] [incumbent [target]|<plan text or file path>] pstack principle-make-operations-idempotent /planning:plan
/planning:draft-goal-condition planning Pick the right autonomy lever and craft a /goal completion condition [--unattended] [intent] pstack autonomous-run /implementation:implement, /session-flow:continue-in-background, /work-items:decompose, /planning:interview
/planning:plan planning Produce a structured implementation plan with an approval gate review, close-out [task description|review|close-out] [--unattended] [--reader guardrails|standard|step-by-step] pstack multi-phase-plan, pstack figure-it-out, pstack principle-exhaust-the-design-space, pstack principle-never-block-on-the-human, pstack feature, Pocock to-spec, Pocock tdd /implementation:implement, /implementation:implement-dispatch, /work-items:decompose
/prototype:explore-directions prototype Throwaway UI variations answering what should this look like vary, combine, controls [vary [<variant>]|combine [<base>]|controls [<variant>]] [--medium <value>] [--level <name>] [scope] pstack prototype /planning:design, /planning:plan, /prototype:pressure-test, /prototype:build-generator
/prototype:pressure-test prototype Throwaway terminal app or shareable HTML demo pressure-testing logic or a data model observe, vendor-fit [observe|vendor-fit] [--medium <value>] [scope] Pocock prototype /planning:design, /planning:plan

7. Decompose

Skill Plugin What it does Actions Arguments Upstream Next
/work-items:decompose work-items Break a plan into vertical-slice work items with dependencies [source] Pocock to-spec, Pocock to-tickets /work-items:work, /work-items:ship

8. Implement

Skill Plugin What it does Actions Arguments Upstream Next
/animation:learn-style animation Measure a traced clip into a style pack and check new scenes against it <rotoscope work dir> <pack dir> [--cuts T,T,..] [--credit TEXT] [--negative SUMMARY.json ...] /animation:produce
/animation:produce animation Brief, boards, approval gate, shots, render, and pack review <production dir> /animation:learn-style
/animation:rotoscope animation Trace, render, measure and fit a 1:1 replica of a reference animation <reference video> <work dir> /animation:learn-style
/architecture:enforce-boundaries architecture Set up a check for each recorded boundary rule and prove it on an injected violation [boundaries findings file] /implementation:implement, /architecture:record-decision, /architecture:improve
/debugging:debug debugging Diagnose broken behavior. Reproduce, hypothesize, instrument, fix with regression test [--unattended] [bug description or observation] pstack tdd, pstack principle-attack-the-premise, pstack feature, pstack runtime-forensics, Pocock diagnosing-bugs /verification:confirm, /review:quality-gate, /implementation:implement, /debugging:analyze-profile, /planning:interview, /bugs:write, /session-flow:workflow
/developer-experience:build-cli developer-experience Build, extend, port or review a team's command-line tools against the CLI contract [review] [what to build or review] /developer-experience:audit-tools, /review:quality-gate
/explainer-video:produce explainer-video Script, narrate, render and check a ManimCE explainer video, frames judged in a fresh context <topic> [output dir] /harness-ops:prerequisites, /source-control:pull-request
/implementation:implement implementation Execute approved plans with TDD, incremental validation, and green commits feature, fix, refactor, replace, config [feature|fix|refactor|replace|config] [task] pstack tdd, pstack autonomous-run, pstack architect, pstack figure-it-out, pstack principle-boundary-discipline, pstack principle-exhaust-the-design-space, pstack principle-laziness-protocol, pstack principle-never-block-on-the-human, pstack feature, pstack visual-parity /testing:plan, /testing:audit, /testing:run-e2e, /review:quality-gate
/implementation:implement-dispatch implementation Orchestrate worker subagents to execute an approved plan [phase] [--unattended] [--items <path>] [--base <ref>] [--wave-cap <N>] pstack autopilot-full, pstack figure-it-out, pstack principle-build-the-lever, pstack orchestrate, pstack feature, Pocock implement-spec /testing:plan, /testing:audit, /testing:run-e2e, /review:quality-gate
/instruction-placement:migrate instruction-placement Move a repository's instruction content to AGENTS.md behind an operator gate plan, apply, cutover-check, remove-shims [plan|apply|cutover-check|remove-shims] [path ...] /instruction-placement:check
/instruction-placement:realign instruction-placement Apply accepted placement findings behind a per-item human gate [finding-id ...]
/overengineering:realign overengineering Execute accepted audit findings down the rollback ladder behind a per-item human gate [finding-id ...] [layer ...] /verification:confirm
/performance:climb performance Keep or revert one measured change at a time until the goal is met [<goal-slug>] pstack benchmark-checklist, pstack hillclimb, pstack perf-issue /performance:verify
/pixel-art:animate pixel-art Author, render, and review pixel-art animation cycles and sprite sheets <character or sprite spec> <cycles> [directions] [engine] /pixel-art:scene
/pixel-art:scene pixel-art Author, render, and review a self-contained HTML pixel-art scene <scene description> [resolution] [duration] [audio file] /pixel-art:animate
/pixel-art:sprite pixel-art Author, render, and review a static pixel-art sprite <what to draw> [size] [palette] [engine] /pixel-art:animate
/pixel-art:tileset pixel-art Author, render, and review an engine tileset or parallax <terrain or set> [engine] [tile size] /pixel-art:scene
/pixel-art:ui pixel-art Author, render, and review pixel-art UI skins, icons, and fonts <window, icons, hud, or font> [engine] /pixel-art:scene
/pixel-art:vfx pixel-art Author, render, and review pixel-art effects and animation cells <effect> [frame count] [engine] /pixel-art:scene
/prototype:build-generator prototype Throwaway generating tool the person tunes, whose saved output goes into the project save-output, add-controls, export [save-output [path/to/config.json]|add-controls [<description>]|export <format>] [<subject>] /verification:confirm
/retro-audio:music retro-audio Render a short chiptune score to a WAV file <mood or score> [chip gameboy|nes|pico-8] /pixel-art:scene
/retro-audio:sfx retro-audio Render one retro sound effect to a WAV file <effect> [preset coin|jump|laser|explosion] /retro-audio:music
/source-control:commit source-control Commit with the resolved convention and surgical staging [message-hint] /source-control:pull-request
/speech:narrate speech Script in, narration.wav and words.json out (kokoro local, or a voice fallback chain) <script or text> [--backend kokoro|elevenlabs] [--out <dir>] [--voice <v>] [--speed <n>] /harness-ops:prerequisites
/user-interface:write-animation user-interface Decide whether an element should animate, then write it in the project's stack [<target>] /user-interface:audit
/user-interface:write-css user-interface Write or review CSS after discovering the project's stack, tokens, targets and conventions [review] [--no-lint] [--show-suppressed] [--target <query>] [what to style or review] /playwright:playwright, /user-interface:setup

9. Test

Skill Plugin What it does Actions Arguments Upstream Next
/app-control:drive app-control Run the app, drive it, record evidence, never judge ready, capture, cleanup, affected [ready|capture --kind <kind>|cleanup|affected] [<app>] [<feature>] [--out <dir>] [--unattended…] /verification:confirm, /debugging:analyze-profile
/app-control:map app-control Generate, prove and keep current a control skill /app-control:drive runs create, check, refresh [create [<app>]|check [<dir>]|refresh [<path>|<feature>] ["<lesson>"]] [--dir <path>] [--unattended] pstack create-verification-skill, pstack maintain-verification-skill /app-control:drive, /app-control:map
/evals:design evals Interview for success criteria and scaffold an eval suite in the consumer repo app, skill, plugin <app|skill <name>|plugin <path>> pstack eval /evals:validate, /evals:plugin-eval
/evals:methodology evals Answer LLM-evaluation design questions from Anthropic's official guidance [question or concept] /evals:design, /evals:plugin-eval
/evals:plugin-eval evals Preflight, price, run, and read a plugin eval suite around the CLI preflight, validate, run, read, ci, init [preflight|validate|run|read <json>|ci|init] [target] pstack eval /evals:validate, /harness-config:unhobble, /evals:design
/evals:validate evals Static FAIL/WARN check of an eval suite's cases and graders, with no model call [eval-dir] /evals:plugin-eval
/mutation-testing:audit mutation-testing Report surviving mutants on the diff or with --exercised, restoration verified, survivors triaged --exercised --full --paths --max --no-suppress --persist-findings --record-mutants --replay-mutants /testing:write, /mutation-testing:principles, /testing:test-value
/mutation-testing:principles mutation-testing Answer mutation-testing questions from the primary literature [question or concept] /mutation-testing:audit, /testing:write
/playwright:playwright playwright Live E2E browser automation with disk-written artifacts [update] [--check|--apply] /source-control:pull-request, /verification:confirm, /playwright:demo-video
/tdd:principles tdd Answer test design questions from authoritative TDD sources [question or concept] Pocock tdd /testing:write
/testing:audit testing Detect tests that cannot fail. Report, gate, or persist [--check] [--strict] [--persist-findings] [--file <path>] [--judge [<base>]] pstack principle-test-behavior-not-implementation /testing:write, /testing:cleanup, /review:fanout, /review:quality-gate
/testing:check-visual-parity testing Baseline screenshots before a change, pixel counts against them after baseline, compare <baseline|compare> [states] pstack visual-parity /implementation:implement, /testing:run-e2e
/testing:cleanup testing Rewrite, quarantine or delete low-value tests behind a mutation gate, or delete cannot-fail ones <<folder>|<repo root>> [--max <n>] [--flaky <test> ...] [--cannot-fail-only] /source-control:pull-request, /code-tidying:audit-dead-code
/testing:diagnose testing Root-cause failing tests, never retry blindly [failure] Pocock diagnosing-bugs /verification:confirm, /testing:write
/testing:plan testing Classify changes by required test type and coverage gaps [range or scope] /testing:write
/testing:run-e2e testing Start the app, drive real flows, capture evidence [worst-case…] [--unattended] [app=<app>] [base=<ref>] [trusted-root=<dir>] [scenario|<target>] pstack create-verification-skill, pstack principle-experience-first /review:quality-gate, /verification:confirm, /testing:diagnose
/testing:test-value testing Where expected values come from and which tests earn their keep pstack principle-test-behavior-not-implementation /testing:audit, /testing:cleanup
/testing:write testing Write and place tests with TDD cadence across ecosystems [task] pstack tdd, pstack principle-test-behavior-not-implementation, Pocock tdd /testing:audit

10. Review

Skill Plugin What it does Actions Arguments Upstream Next
/code-tidying:audit-comment-residue code-tidying Classify code comments for history narration and session-reference residue [audit] [--added-since <base>] [target] /code-tidying:remove-comments, /code-tidying:tidy, /debugging:debug, /implementation:implement
/code-tidying:batch-simplify code-tidying Batch-run simplification across changed files, or a whole repository, by ecosystem time-window, branch, repo [time-window|branch|repo] [--unattended] [path...] [docs] [override] [in-place[=commit]] /verification:confirm
/code-tidying:remove-comments code-tidying Remove comments that do not earn their keep, under a posture; audit comment residue read-only audit, report, safe [audit|report|safe] [none|necessary|explanatory] [override] [--pipeline] [--base <ref>] [target] pstack no-comments, pstack principle-attack-the-premise /source-control:commit, /review:fanout, /implementation:implement, /debugging:debug, /architecture:improve, /architecture:record-decision
/code-tidying:tidy-diff code-tidying Tidy one change's own lines before its commit, scoped by a file-and-line list <lines-file> [--turns <n>] /source-control:commit
/harness-ops:audit-mod-posture harness-ops Inventory installed and built-in mods with their hooks and calls, and the guard state [--builtin-json <inventory-report>] [--project <dir>] [--surface desktop] /harness-config:audit-permission-state
/mcp-tools:audit mcp-tools Audit MCP tool definitions against design quality criteria [path]
/mcp-tools:audit-posture mcp-tools Inventory configured MCP servers and flag floating versions and other supply-chain risks [--config <file> ...] /harness-config:audit
/plugin-quality:audit plugin-quality Behavioral audit of a plugin component ending in a maintainer work item session, arm [<plugin>[:<component>]...|session|arm] /work-items:work
/review:code-review review Org CI code-review lane command for a GitHub pull request
/review:council review Configurable reviewer panel chosen per change, with optional fix rounds [--yes] [--unattended] /review:fanout, /verification:confirm
/review:explain-change review Change digest for a pull request, markdown record plus an interactive view pr-number [pr-number|this branch] [--event ready] [--policy off|offer|always] [--quiz] /review:quality-gate, /review:reviewer-media
/review:fanout review Fan review out across every reviewer surface into one ranked report run-everything, fix [run-everything|fix] [--yes] [--unattended] pstack autopilot-full, pstack interrogate /review:fanout, /verification:confirm, /review:council
/review:merge-risk review Checked merge-risk rating of one pull request with a demote-only flag pr-number, pr-url [pr-number|pr-url|this branch] [--dir <dir>] /review:explain-change
/review:quality-gate review Single-lens review checkpoint routed to the matching reviewer self, code, code-design, architecture, security, spec, close-out, downstream, pr, criteria, slice [self|code|code-design|architecture|security|spec|close-out|downstream|pr|criteria|slice…] pstack blast-radius, pstack interrogate /verification:confirm
/review:reviewer-media review Narrated walkthrough video for a pull request's reviewers, placed in its description <pr> [--handoff] [--policy <policy>] [--blast-radius <level>] [--out <dir>] /source-control:pull-request
/review:security-review review Org CI security-review lane command for a GitHub pull request
/skill-quality:check skill-quality Static QA gate for skill frontmatter, caps, and evals check, validate-evals, listing-budget, measure-invocation [check|validate-evals|listing-budget|measure-invocation] [<skill-or-root> ...] /playbooks:skill-authoring, /verification:confirm
/toolchain:ratchet toolchain Hold a count in CI with a checked-in ceiling a human merges [<counter, rule or claim>] pstack correct /source-control:pull-request
/user-interface:audit user-interface Review interface code read-only by concern lane, with severity and evidence per finding diff, tree, launch [diff|tree|launch] [<concern>] [opportunities|performance|removal|cleanup] [--fix] [<target>] /user-interface:write-animation

11. Verify outcome

Skill Plugin What it does Actions Arguments Upstream Next
/performance:protect performance Hold a verified counter win with a CI ceiling a human merges [<counter or claim>] /toolchain:ratchet, /source-control:pull-request
/performance:snapshot performance Capture a snapshot only from a host proven measurable baseline, post [baseline|post] [<target>] pstack hillclimb /implementation:implement, /performance:snapshot, /performance:climb, /performance:verify
/performance:verify performance Re-derive the result in fresh context and report it honestly [<claim or target>] pstack principle-explain-the-number, pstack benchmark-checklist /performance:protect, /source-control:pull-request, /performance:target, /debugging:debug
/playwright:demo-video playwright Replay a web flow, render a produced demo video, quality-control it from frames, post it to the PR <url or flow to demo> [--style produced|plain] /playwright:playwright, /source-control:pull-request
/toolchain:check toolchain Build, test, and lint changed files across detected ecosystems [ecosystem] /verification:confirm, /testing:diagnose
/toolchain:lint toolchain Polyglot lint and format checks without a full build [ecosystem] [--fix|--code-fix] [--yes] [--dry-run] [--all-files]
/verification:confirm verification Prove the change achieved its intended outcome with evidence outcome, fix, refactor [outcome|fix|refactor] [--unattended] [ecosystem|all] pstack create-verification-skill, pstack principle-prove-it-works, pstack autopilot-full, pstack multi-phase-plan, pstack principle-laziness-protocol /source-control:pull-request, /verification:measure, /verification:confirm, /app-control:map, /corrections:codify
/verification:measure verification Verify an improvement claim against a pre-change baseline performance, metrics [performance|metrics] [baseline|compare] /verification:confirm, /performance:snapshot

12. Retrospective

Skill Plugin What it does Actions Arguments Upstream Next
/corrections:codify corrections Count a recurring mistake, pick the strongest rung and prove the check findings, reviews, issues, note, preferences, group, count-checks <<description> [#PR]|findings <file>|reviews|issues|note <text>|preferences|group|count-checks…> pstack automate-me, pstack principle-encode-lessons-in-structure, pstack correct /harness-config:audit-automation-gaps, /architecture:improve, /corrections:codify, /implementation:implement
/session-flow:audit-friction session-flow Mine sessions for permission and autonomy friction and end in one decision brief [remeasure] [--days <n>] [--project <text>] [--source <dir>] [--verify consequential|full|probes] /session-flow:audit-friction
/session-flow:audit-sessions session-flow Cross-session audit of transcripts with routed, never-applied findings sweep, pr, trends [sweep|pr|trends…] [--scope machine|project] [--since <day>] [--until <day>] [--write-report] pstack automate-me /corrections:codify, /session-flow:audit-friction
/session-flow:retro session-flow Structured session retrospective with codified learnings session, quick [session [--pattern '<regex>']|quick] [--unattended] Pocock retro, Pocock chief-of-staff /session-flow:audit-sessions
/session-flow:running-retro session-flow In-flight retro checkpoint appended to a running ledger topic, arm [topic|arm] Pocock retro /session-flow:retro

Ship: PR lifecycle (after stage 11)

Skill Plugin What it does Actions Arguments Upstream Next
/source-control:pull-request source-control Full PR lifecycle. Prep, create, monitor CI, address reviews, merge prep, create, ready, monitor, comments, ci, merge, reassess, evidence, status, full [prep|create|ready|monitor|comments|ci|merge|reassess|evidence|status|full|…] [--unattended] [args] pstack opening-a-pr, pstack shipping, pstack babysit, pstack multi-phase-plan, Pocock pr, Pocock resolving-merge-conflicts /session-flow:retro, /implementation:implement, /review:reviewer-media

Anytime / cross-cutting

Skill Plugin What it does Actions Arguments Upstream Next
/adhd:clarify adhd Reshape a dense, decision-heavy message into clear one-decision-at-a-time chunks, losing nothing [--medium <medium>] [artifact to clarify]
/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, commit messages, PR bodies and interface copy audit, fix [audit|fix] [target|user-scope [memory]|commit [rev]|pr-body|copy <files>] pstack unslop
/architecture:principles architecture Define the module-design terms architecture findings use [term or question] /architecture:improve
/architecture:record-decision architecture Record an architecture decision in the repository's existing ADR convention [audit] [decision and its rationale, or a path to a file holding them] /planning:plan
/attribution:audit attribution Find prose copied from external sources and convert it into pointers audit, fix, sweep [audit|fix|sweep] [target]
/code-metrics:audit code-metrics Lines, complexity and typed-code numbers beside cited references, no verdict size, complexity, type-debt <size|complexity|type-debt> [--json] [--all] [--base <ref>] [<path>...] pstack interrogate /work-items:track, /verification:measure, /code-metrics:audit-coverage, /code-metrics:principles
/code-metrics:audit-coverage code-metrics Coverage and CRAP read from build artifacts, no verdict [--json] [--all] [--base <ref>] [--artifacts <path>]... [<path>...] /verification:measure, /mutation-testing:audit, /code-metrics:principles
/code-metrics:audit-duplication code-metrics Clone groups minus the replication the repo declares, no verdict [--json] [--keep] [--all] [--base <ref>] [--registry <file>] [<path>...] /verification:measure, /code-metrics:principles
/code-metrics:audit-suppressions code-metrics Suppressions with their rule ids and reasons, no verdict [--json|--findings [--memory-dir <dir>]] [--all] [--base <ref>] [<path>...] pstack no-comments /toolchain:ratchet, /review:fanout, /code-metrics:principles
/code-metrics:principles code-metrics What each code measure can and cannot tell you [question or concept] /code-metrics:audit, /code-metrics:setup
/code-tidying:audit-dead-code code-tidying Whole-repo dead-code hunt across four labeled lanes with adjudicated candidates [--max N] [--lane knip|vulture|gopls|grep] [target] /code-tidying:tidy
/code-tidying:tidy code-tidying Proactively hunt one lane for safe structural tidyings and ship a structure-only PR dry-run, self-update, help [<lane>|<glob>...|dry-run [<lane>|<glob>...]|self-update|help] [override] [in-place[=commit]] /source-control:babysit-prs
/codebase-health:audit codebase-health Audit for drift between docs, config, code, and architecture via verified findings [scope] [--fix] [--medium <value>] [--docs-only|--code-only|--config-only|--arch-only] /implementation:implement, /verification:confirm
/computer-use:diagnose computer-use Resolve computer-use capture, input, and screenshot symptoms to a cause screenshots, failures, quirks [screenshots|failures|quirks]
/context-budget:audit context-budget Measure the startup context payload per item and ledger every lever's real delta fix [--full-sweep|--tools T1,T2|--ledger|--measure-only|fix] [--unattended]
/coupling:reduce coupling Scan for change-transmitting coupling, apply safe reductions in a budgeted batch, route the rest dry-run, change, status, help [<scope>|dry-run [<scope>]|change [apply] <old> <new> [<scope>]|status|help] pstack principle-minimize-reader-load /verification:confirm
/developer-experience:audit-tools developer-experience Inventory a repository's developer tooling and report findings, read-only by default [path or repo ...] /developer-experience:build-cli, /overengineering:justify
/discipline:be-ambitious discipline Raise a request before work and wait, or deepen finished work and redo it with an outside check [depth] [request|feedback] /planning:interview
/discipline:do-your-research discipline Re-anchor research discipline, then audit and correct the current work tiered, full [tiered|full] /discovery:research
/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 [target]
/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 [fan-out] [scope]
/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 [lever-check <block summary>] pstack principle-build-the-lever /implementation:implement, /harness-config:audit-automation-gaps
/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 Pocock wait-what
/disk-hygiene:audit disk-hygiene Scan a directory tree for stale leftovers and report the evidence, read-only [--max-depth <N>] [--sizes-only] [--policy <file>] <target-directory> /disk-hygiene:clean
/disk-hygiene:clean disk-hygiene Audit a directory tree for stale leftovers and remove validated paths [--execute] [--deep] [--max-depth <N>] [--sizes-only] [--policy <file>] [--quiet] <target-directory>
/docs-hygiene:audit-derivability docs-hygiene Judge whether a doc earns its existence or should become a pointer target, audit, sweep [target|audit [target]|sweep <dir>]
/docs-hygiene:audit-encapsulation docs-hygiene Find external citations reaching into a skill's private surfaces detect, sweep, file-issues [detect|sweep|fix <file>:<line>|file-issues]
/docs-hygiene:audit-noise docs-hygiene Classify markdown for citations, ghost refs, meta-commentary, plan/conversational/tracker residue [audit] [target] [--persist-findings] /review:fanout, /docs-hygiene:audit-derivability, /docs-hygiene:compress
/docs-hygiene:audit-progressive-disclosure docs-hygiene Grade instruction files for split opportunities and hub/spoke disclosure defects [audit] [target] /docs-hygiene:write-for-agents, /docs-hygiene:compress, /docs-hygiene:rename-references
/docs-hygiene:compress docs-hygiene Tighten markdown by dropping flavor while preserving every directive audit, compare, target [audit [target]|compare <ORIG_DIR> <NEW_DIR> [REASONS]|target] [--force] [--keep-snapshot]
/docs-hygiene:extract-ssot docs-hygiene Deduplicate repeated prose into one named source of truth identify, verify, plan, execute, batch, unwind [identify|verify|plan|execute|batch|unwind] [<cluster>] [--fix] [--dry-run] [--yes]
/docs-hygiene:rename-references docs-hygiene Sweep stale references after renames, including forms grep misses audit, preview, blocklist [audit|preview|blocklist] [<old> [to <new>]] [--container|--identifier|--include-historical…] /docs-naming:audit-file-names
/docs-hygiene:write-for-agents docs-hygiene Authoring-time doctrine for agent-consumed markdown [<file or section being written>]
/docs-hygiene:write-for-humans docs-hygiene Authoring-time doctrine for human-read documentation [<doc or section being written>] pstack technical-writing
/docs-naming:audit-file-names docs-naming Inventory a doc tree's file names and plan the renames with their references [audit] [root ...] /docs-naming:realign-file-names
/docs-naming:realign-file-names docs-naming Apply a file-name rename plan, one acceptance per file [FN-xxxxxxxx ...]
/education:explain education Explain any concept or the last response in genuinely plain words [thing to explain] pstack teach /education:teach, /education:illustrate
/education:illustrate education Visual explainer for a concept or codebase topic, record plus interactive page [topic] [zero-knowledge] [ste] [markdown] [medium] /education:teach, /education:quiz-me
/education:quiz-me education Generate a post-change report with a quiz verifying you absorbed the work [recall <query>]
/education:teach education Multi-session learning coach for general topics or repo-grounded concepts topic, codebase, mission, glossary, resources, explain, primer, exercise, assess, resume, status <topic|codebase|mission|glossary|resources|explain|primer|exercise|assess|resume|status> [args] pstack teach, Pocock teach
/fleet:reach fleet Reach another fleet lane (WSL or Windows, here or remote) to run, prompt, query or message [relay]
/github:advise github Design and set up GitHub settings and admin areas grounded in live gh state [--unattended] [topic] [--apply]
/github:audit github Read-only audit of GitHub org and repo settings, drift, and cost signals [area ...] [--apply]
/harness-config:audit harness-config Audit settings, hooks, permissions, and MCP config for drift against current official docs [--fix] [permissions|mcp|hooks|plugins|issues|effort-pins|all] /harness-config:audit-permission-grants, /harness-config:audit-automation-gaps, /harness-config:audit-pass
/harness-config:audit-automation-gaps harness-config Audit the repo's automation landscape for hook, MCP, skill, and subagent gaps worth adding [--recommend-only] [--implement] [hooks|mcp|skills|subagents|scheduled|all] /planning:plan, /work-items:decompose, /overengineering:audit, /work-items:track
/harness-config:audit-instructions harness-config Find instructions current models no longer need across CLAUDE.md, AGENTS.md, rules, and skill bodies [scope] [--target-model <version>] [--opinion] [--persist-findings] [--unattended] /harness-config:unhobble, /harness-memory:audit, /instruction-placement:migrate, /harness-config:audit-prompting-postures
/harness-config:audit-pass harness-config Run one coordinated, resumable audit pass over a repo with a single human gate [target] [--fix] [--opinion] [--postures] [--lanes <list>] [--resume] [--report-to <dir>]
/harness-config:audit-permission-grants harness-config Audit permission grants for portability and auto-mode durability frontmatter, settings, plugins, all [frontmatter|settings|plugins|all] [--unattended] /harness-config:audit-permission-state, /harness-config:audit
/harness-config:audit-permission-state harness-config Report the permission rules actually in effect and what auto mode drops [--unattended] [--scopes] [--entry-diff] [--lint] [--managed] [--block] [--oracle] [--critique]
/harness-config:audit-prompting-postures harness-config Find posture guidance the prompting guide says a component needs but does not carry skills, agents, hooks, output-styles, claude-md, rules, all [skills|agents|hooks|output-styles|claude-md|rules|all]
/harness-config:draft-auto-mode-rules harness-config Interview and draft a paste-ready autoMode block, never writing settings environment, allow [environment|allow|soft_deny|hard_deny] [--unattended]
/harness-config:unhobble harness-config Strip instructions to a bare baseline, log real stumbles, re-add only what evidence earns snapshot, bare, observe, readd, watch, status, decide [snapshot|bare|observe|readd|watch|status|decide]
/harness-memory:audit harness-memory Audit CLAUDE.md, a root AGENTS.md, rules, and auto-memory against the official-docs checklist audit, fix, update, report [audit|fix|update|report]
/harness-memory:stateless harness-memory Inspect, disable, or purge Claude Code's per-repo auto memory status, disable, purge [status|disable|purge] [--unattended]
/harness-ops:audit-plugin-landscape harness-ops Report plugin and marketplace staleness, overlaps and alternatives; recommend only [--plugins-dir <path>]
/harness-ops:behavior-probes harness-ops Run data-driven live probes of Claude Code permission and platform behavior and record the outcomes validate, dry-run, live, recheck [validate|dry-run|live [--case <id>]...|recheck <range>|add <area>/<name>] /harness-ops:changelog
/harness-ops:changelog harness-ops Ingest a Claude Code release changelog and integrate its changes into the repo fetch, diff, status, apply <fetch|diff|status|apply> [--ci] [vA..vB|vX|text] /harness-ops:audit-native-overlap, /work-items:triage, /harness-ops:inventory, /harness-ops:behavior-probes
/harness-ops:known-issues harness-ops Look up and track known Claude product issues, health, and workarounds status, search, check-all, scan, list, quality, create <status|search|check-all|scan|list|quality|create> [args]
/improvement:find improvement Rank evidence-cited improvement candidates across dimensions; execution goes to the pipeline [target] [--small|--medium|--large] [--unattended] [repo-path] /improvement:improve, /planning:interview
/improvement:improve improvement Find one gap in a target against a stated standard and ship the fix as one PR [dry-run] [target ...] [--focus <lens>] [--unattended] /source-control:pull-request, /planning:plan
/instruction-placement:audit instruction-placement Find instruction content on the wrong surface and propose validated destinations core, expanded [core|expanded] [path ...] /instruction-placement:realign, /instruction-placement:migrate
/instruction-placement:check instruction-placement Gate that every path-scoped rule glob resolves and the rules index is current [--file <index-path>] [--breadth-max <pct>] /instruction-placement:realign
/instruction-placement:delta instruction-placement Report only what moved since the last placement audit [--since <ISO date>] [--noise-budget <n>] /instruction-placement:realign
/multi-agent:audit-defaults multi-agent Recheck the routing defaults against their upstream sources fanout, repo [<role>|fanout|repo] /multi-agent:route
/overengineering:audit overengineering Audit the enforcement surface for mechanisms no longer earning their carry cost [--unattended] [--medium <value>] [layer ...] /overengineering:realign
/overengineering:delta overengineering Re-run the enforcement-surface audit and report only what moved since the last run [--unattended] [layer ...] /overengineering:realign
/overengineering:justify overengineering Make one artifact, or one proposal before it is built, justify its existence on evidence path, proposal [path|path#heading|kind:identifier|proposal] /overengineering:realign, /planning:plan
/playbooks:boris playbooks Boris Cherny's Claude Code workflow tips across 131 sections (130 upstream, L1 local)
/playbooks:fable-5 playbooks Fable 5.2's operating doctrine loaded as standing session instructions full [full|<chapter>] pstack principle-never-block-on-the-human, pstack in-your-own-words
/playbooks:repo-sweep playbooks Run the hygiene skill catalog through one repo, one PR, one commit per step plan, next, review [plan|next|review] /source-control:pull-request
/playbooks:skill-authoring playbooks Anthropic's internal skill-authoring playbook and patterns pstack authoring-a-skill /skill-quality:check
/playgrounds:use playgrounds Route playground requests to the first-party plugin, or install it, with recipes and guidance [what to explore]
/repo-hygiene:clean repo-hygiene Clean caches, build artifacts, stale branches, and stashes per repo scan, caches, build, git, stash, tree, all, aliases [scan|caches|build|git|stash|tree|all|<tier>-batch|aliases…] /source-control:worktree, /repo-fleet-hygiene:audit, /disk-hygiene:clean
/session-flow:recommend-skills session-flow Lay out the skills that fit this moment as a ranked, nothing-hidden menu topic-slug, question [topic-slug|question] pstack poteto-help /session-flow:workflow
/session-flow:workflow session-flow Navigate the staged dev workflow and suggest the next stage steps, pre-pr, wrap-up, philosophy, spec-first, continue [steps|pre-pr|wrap-up|philosophy|spec-first|continue [auto]] pstack poteto-mode, pstack poteto-help, Pocock ask-matt
/user-experience:evaluate user-experience Plan and run UX evaluations, fair-choice checks and measure selection, evidence-labeled [design, flow, app or measurement question]
/user-experience:plan-user-research user-experience Plan user research and write its instruments, evidence-labeled [research question or app context] /user-experience:synthesize
/user-experience:shape user-experience Detect the app's stage and the project's own evidence, then chain the UX jobs in order [the app or the UX question] /user-interface:design, /user-experience:setup
/user-experience:synthesize user-experience Synthesize research data into labeled themes, insights, needs, personas or jobs [research data paths or app context] /user-experience:structure
/user-interface:glossary user-interface Name the design or front-end term for a described effect, detail or behavior [<description>] /user-interface:write-animation, /user-interface:write-css
/visualization:present visualization Slide deck through the claude.ai Slides Artifact type, outline markdown as the record [topic or source] [<medium>] /visualization:visualize
/visualization:visualize visualization Pick the best visual form for what is in the conversation and render it [<medium>] [--unattended] Pocock pr /visualization:present
/wizard:generate wizard Author a hardened interactive bash wizard for human-only setup, credential, and cutover steps <procedure to wizardize> Pocock wizard /wizard:unattended
/wizard:unattended wizard Author an unattended script a human launches once for a privilege or policy boundary <scriptable procedure a human must launch> /wizard:generate
/work-items:measure-delivery work-items Report the five DORA delivery metrics as a team trend from default-branch history [--since YYYY-MM-DD] [--until YYYY-MM-DD] [--bucket week|month] [--branch <ref>] [--merge-risk] /work-items:track
/work-items:scan-todos work-items Sweep source comments for TODO and FIXME markers, resolve or file each [--path <dir>] [--work] /work-items:triage
/work-items:ship work-items Route and run a spec container's macro journey. Status, execution shape, next step status, run [status|run] [#<container-id>|<topic-slug>] Pocock implement-spec /work-items:work, /work-items:ship
/work-items:track work-items Backlog CRUD through the bound tracker. Add, list, close, stats stats, list, add, start, done, due, recheck, search, audit, publish-plan, link, view, edit, changes [stats|list|add|start|done|due|recheck|search|audit|publish-plan|link|view|edit|changes…] [args] /work-items:work, /work-items:triage
/work-items:triage work-items Evaluate raw intake through the verified-to-eligible state machine [sweep] [--config-ref <ref>] [<number>] Pocock triage /work-items:work, /planning:interview, /debugging:debug, /work-items:decompose
/work-items:work work-items Auto-select one work item and execute it end-to-end [<item-id>|--issue <n> --ci] pstack in-your-own-words /source-control:babysit-prs, /work-items:work
/writing:be-concise writing Reshape prose for a scanning reader. Bottom line first, about half the words [target]

Session lifecycle

Skill Plugin What it does Actions Arguments Upstream Next
/multi-agent:assess multi-agent Decide workflow, subagents or one context for a task <task description> pstack autopilot-full /multi-agent:route, /session-flow:orchestrate
/multi-agent:route multi-agent Resolve model and effort per agent role for a multi-agent run all <<role>|all> [workload] [session=<alias>] pstack autopilot-full, pstack setup-pstack /multi-agent:audit-defaults
/plugin-profiles:profile plugin-profiles Switch, check or set up the plugin profile a folder or session loads apply, launch, status, clear, check, setup [apply <name>|launch [<name>]|status|clear|check|setup] /harness-ops:audit-skill-visibility, /session-flow:handoff
/session-flow:clean-stop session-flow Make everything durable before the machine goes away [--unattended]
/session-flow:continue-in-background session-flow Delegate the task to a fresh background agent now [--unattended] [--cloud] [file|prompt] [topic] [purpose...]
/session-flow:find-handoff session-flow Recover a lost handoff or resume prompt after /clear pstack session-pickup
/session-flow:handoff session-flow Write a mid-session save-point for clear-and-resume file, prompt [file|prompt] [--unattended] [topic] [purpose...] pstack pause-safely /session-flow:find-handoff, /session-flow:reanchor
/session-flow:keep-going session-flow Recover after an interruption and continue where work stood pstack orchestrate, pstack session-pickup /session-flow:workflow
/session-flow:orchestrate session-flow Arm the session with proactive-orchestration imperatives handoff, worker, coordinator [<task>|handoff [compact]|worker [compact]|coordinator [<area>]] pstack autopilot-full, pstack poteto-mode, pstack orchestrate, Pocock chief-of-staff /planning:draft-goal-condition
/session-flow:orient session-flow Read-only situation report from durable and off-thread state catch-up, zoom-out [catch-up|zoom-out [--days <N>]] [--unattended] pstack ketchup
/session-flow:reanchor session-flow Verify working assumptions are still true before building on them
/session-flow:recall session-flow Topic history from past transcripts, reverts, closed PRs and tracker items <topic> pstack recall /session-flow:orient
/session-flow:reconcile session-flow Retire finished off-thread work and square the task ledger
/session-flow:tidy-work session-flow Report, normalize, and clean stale .work memory items report, normalize, clean [report|normalize|clean] [--days N] [--offline] [--memory-dir DIR] /session-flow:clean-stop
/source-control:worktree source-control Create, inspect, and clean git worktrees for parallel sessions create, status, cleanup, audit [create|status|cleanup|audit] [args] pstack worktree-cleanup

Operator cadence

Skill Plugin Cadence What it does Actions Arguments Upstream Next
/bugs:scan bugs daily Proactively hunt resting code for unobserved bugs, verify adversarially, report read-only [<path|feature|diff>] [--lane <name>] [--track] [--dry-run] [--unattended] /implementation:implement, /work-items:track, /review:security-review, /debugging:debug
/harness-ops:audit-install-state harness-ops weekly Audit a Claude Code install directory. What is there, what the product manages, what is stale [--unattended] [root] /disk-hygiene:clean
/harness-ops:audit-native-overlap harness-ops weekly Map native Claude Code surfaces against this repo's components and record owner-directed verdicts report, apply, dismiss [report|apply <plugin>|dismiss <native> <plugin:name> <reason>] [--store <p>] [--inventory <p>]
/harness-ops:audit-performance harness-ops continuous Capture slowness evidence while slow. Version, sweep health, tree walk, sessions, fleet, fan-out [--unattended] [--root <path>] [--session-id <id>] [--note <fact>]
/harness-ops:audit-skill-visibility harness-ops weekly Which skills the model can actually see, which are starved, and which are unobservable [--unattended] [--installed [dir]] [--plugins-root <dir>] [--render markdown|json]
/harness-ops:inventory harness-ops weekly Enumerate every command, skill, agent, and plugin component this machine can invoke [--builtin|--plugins|--bundled|--agents|--tools|--hooks|--docs] [--marketplace <m>] [--diff <f>] /harness-ops:audit-native-overlap
/harness-ops:lanes harness-ops daily Start, restart, stop, and check loop lanes as named background sessions start, restart, status, stop, run-once, print-schedule, consume-restarts [start|restart|status|stop|run-once|print-schedule|consume-restarts] [lane...]
/harness-ops:machine-profile harness-ops weekly Discover, store and diff machine facts and per-tree identity domains. Read-only by default. profile, diff, explain, apply [profile|diff|explain <key>|apply --option <key>] /harness-ops:prerequisites, /harness-ops:plugins-audit
/harness-ops:observability harness-ops weekly Report on locally captured telemetry. Token burn, cost, hook latency, per-session activity week, session, day, month, all, clean, latency, compare, items, hooks [week|session|day|month|since:DATE|all|clean|latency|compare|items|hooks] [--write] [--days N]
/harness-ops:plugins harness-ops weekly Bring the machine's plugin fleet current. Refresh, update, install per policy sync, converge [sync|converge] [<marketplace>|all] [--allow-downgrade] [--update-other-paths] /harness-ops:prerequisites
/harness-ops:plugins-audit harness-ops weekly Predict what plugin sync would change, read-only all [<marketplace>|all] /harness-ops:plugins
/harness-ops:prerequisites harness-ops weekly Report external tools the enabled fleet declares missing. Never installs. /harness-ops:machine-profile, /harness-ops:plugins-audit
/repo-fleet-hygiene:apply repo-fleet-hygiene weekly Execute a fleet action plan behind one confirmation gate --plan-file <path> [--apply] [--yes] [--remote-branches]
/repo-fleet-hygiene:audit repo-fleet-hygiene weekly Discover a repository fleet and coordinate read-only evidence handoffs [[<dir>]... [--root <dir>]... [--repo <dir>]... [--config <file>] [--detail]|--apply-plan <path>]
/repo-fleet-hygiene:sync repo-fleet-hygiene weekly Fast-forward canonical checkouts to the remote default branch [<dir>...] [--root <dir>] [--repo <dir>] [--skip <name>] [--config <file>] [--apply] [--yes] /repo-fleet-hygiene:audit
/source-control:babysit-loop source-control continuous Run one repo's PR queue as a standing merge lane <owner/repo> [safe|worker|autopilot] [--drain] [--strip-do-not-merge] /work-items:attend-queue
/source-control:babysit-prs source-control continuous Tiered fleet pass advancing your open PRs worker, autopilot, help [worker|autopilot|help] [--unattended] [owner/repo|#n|owner/repo#n] pstack babysit
/work-items:attend-queue work-items daily Drive escalations to resolution; list held PRs and post-merge failures; hand intake to triage /work-items:work-loop
/work-items:morning-brief work-items daily Print the operator's read-only morning view. Queues, merge-ready PRs, parked decisions [precision] [--repo owner/name] [--queue-labels A,B,C] [--stale-hours N] [--stuck-days N]
/work-items:work-loop work-items continuous Drain the backlog as a self-paced autonomous loop [<owner/repo>] [--drain|--standing] [--single-pass] [--shard <i>/<n>] [--ordering newest-first…] pstack autopilot-full, Pocock software-factory /work-items:attend-queue

Review and pull-request skill map

Skill Job Who it is for Posts to the PR
/review:quality-gate Single-lens review checkpoint of a change before it is ready author, reviewer no
/review:fanout Every reviewer at once, merged into one ranked report on disk author, reviewer no
/review:council Configured panel of reviewers, one ranked report, optional fix rounds author, reviewer no
/review:explain-change Digest of one PR: why, before and after, risk map, where to focus reviewer no
/review:merge-risk Checked merge-risk rating of one PR, with a demote-only flag reviewer no
/review:reviewer-media Decides whether a PR needs a walkthrough video and builds it from the digest author, CI lane through another skill
/review:code-review CI lane: correctness and maintainability findings on one PR CI lane yes
/review:security-review CI lane: logic, trust-boundary and Actions security findings on one PR CI lane yes
/source-control:pull-request Whole PR lifecycle: prep, create, ready, monitor, merge, evidence, media author yes
/source-control:babysit-prs Tiered fleet pass advancing open PRs; resolves threads and merges only at worker and autopilot author yes
/source-control:babysit-loop One repository's PR queue run as a standing merge lane CI lane through another skill
/playwright:demo-video Produced demo video of a web UI change, quality-checked from frames author yes
/explainer-video:produce Renders a narrated video from a topic or a storyboard author no