Skills Reference

Skills Catalog

All 23 skills shipped through 8-habit-ai-dev v2.18.1. Skills are read-only guidance — they tell Claude how to approach a task, they do not modify files themselves.

Tip

Not sure which skill to use? Type /using-8-habits for an interactive decision tree, or see the quick-select matrix below.

7-Step Workflow Skills

#	Skill	Habit	When to use
0	/research	H5	Problem space is unclear — investigate before specifying
1	/requirements	H2	Any new feature or non-trivial change
2	/design	H8	Architecture decisions — human decides
3	/breakdown	H3	Decompose work into atomic tasks
4	/build-brief	H5	Before coding each task — read existing code first
5	/review-ai	H4	Always before committing AI-generated code
6	/deploy-guide	H1	Staging-first deploys with rollback plan
7	/monitor-setup	H7	Observability after deploy

Assessment Skills

Run at any point in the workflow for deeper analysis.

/cross-verify {#cross-verify}

17-question 8-Habit cross-verification checklist. Run after planning and before committing to implementation. Produces a dimension summary across all 8 habits with confidence levels.

• Habit: H1-H8 (all)
• When to use: Before ExitPlanMode, before PR creation
• Source: skills/cross-verify/SKILL.md

/whole-person-check {#whole-person-check}

Assess a feature/component/PR against Covey's 4 dimensions: Body (discipline), Mind (vision), Heart (passion), Spirit (conscience). Scores 1-5 per dimension.

• Habit: H8 Find Your Voice
• When to use: After /review-ai when you want a balance check
• Source: skills/whole-person-check/SKILL.md

/security-check {#security-check}

Focused security review — secrets, injection, auth, dependencies, OWASP Top 10. A dedicated security lens separate from general code review.

• Habit: H1 Be Proactive
• When to use: Any change touching user input, auth, data handling, or dependencies
• Source: skills/security-check/SKILL.md

/scrutinize {#scrutinize}

Outsider-perspective end-to-end review — questions intent first ("is there a simpler way?"), then traces the actual code path (not just the diff) to verify the change does what it claims. Pairs with /review-ai (scope-question vs diff-local). Operating Rules hardened in v2.18.1 (ADR-017) with MUST/NEVER + Why blocks.

• Habit: H5 Understand First + H8 Find Your Voice
• When to use: Before committing to an implementation, or before merging a PR alongside /review-ai
• Source: skills/scrutinize/SKILL.md

/consistency-check {#consistency-check}

Cross-artifact consistency analyzer — runs 5 detection passes (Coverage, Drift, Ambiguity, Underspec, Inconsistency) across persisted spec artifacts (PRD ↔ design ↔ tasks). Read-only — reports drift, coverage gaps, contradictions. Inspired by github/spec-kit /analyze.

• Habit: H5 Understand First + H1 Be Proactive
• When to use: After /requirements, /design, /breakdown have been run with --persist <slug> to catch drift before code
• Source: skills/consistency-check/SKILL.md

/reflect {#reflect}

Post-task micro-retrospective — 6 questions, 5 minutes. Captures lessons to ~/.claude/lessons/ for future retrieval by /research and /build-brief. Includes skill-effectiveness signal (Q6) feeding SKILL-EFFECTIVENESS.md.

• Habit: H7 Sharpen the Saw
• When to use: After completing a task or workflow
• Source: skills/reflect/SKILL.md

/workflow {#workflow}

Interactive guided walkthrough of the 7-step workflow. Prompts at each step to invoke or skip.

• Habit: All
• When to use: Starting a new feature when you're unsure which step comes next
• Source: skills/workflow/SKILL.md

Debug Discipline Skills

Active bug investigation and post-fix engineering record. Ported from prior-art (mattpocock/skills, 9arm-skills) per ADR-014 + ADR-015.

/diagnose {#diagnose}

6-phase active bug investigation methodology — feedback-loop → reproduce → hypothesise → instrument → fix-with-regression-test → cleanup. Closes the gap between /research (too broad) and /post-mortem (too late). Phase 6 hardened in v2.18.1 (ADR-017) with MUST re-run Phase 1 feedback loop, citing Anthropic pptx.

• Habit: H1 Be Proactive + H5 Understand First
• When to use: When a bug needs to be fixed and the cause isn't obvious from stack trace; hands off to /post-mortem once fix lands
• Source: skills/diagnose/SKILL.md

/post-mortem {#post-mortem}

Engineering RCA writeup — canonical record of a fixed bug (root cause, mechanism, fix, validation, how it slipped through). Refuses to draft without 4 inputs (reliable repro, known cause, identified fix, validated outcome).

• Habit: H4 Win-Win + H7 Sharpen the Saw
• When to use: AFTER a debug session lands a validated fix, BEFORE closing the ticket
• Source: skills/post-mortem/SKILL.md

Spec & Communication Skills

Helpers for project orientation and audience reshape — orthogonal to the 7-step workflow.

/save-spec {#save-spec}

Scaffolds a project-root SPEC.md digest following the spec-digest-pattern archetype (project orientation hub). User-invoked. Generator-only Phase 1; refuses to overwrite an existing SPEC.md. Distinct from per-feature /requirements --persist <slug> mode.

• Habit: H2 Begin with End in Mind + H5 Understand First
• When to use: When starting on an unfamiliar repo or after a /clear flushing pain — gives Claude a 4-section orientation hub
• Source: skills/save-spec/SKILL.md

/management-talk {#management-talk}

Channel-aware audience reshape — converts engineer-to-engineer content into leadership-channel-shaped form (JIRA / Slack / standup / email / meeting). Strips function/file/SHA noise but keeps JIRA keys, PR numbers, workload names. Inspired by 9arm-skills (v2.17.0).

• Habit: H4 Win-Win + H6 Synergize
• When to use: When VPs, directors, PMs, or release managers need a status — and your raw engineering notes are too dense
• Source: skills/management-talk/SKILL.md

Meta & Onboarding Skills

Learn the plugin and adapt it to your style.

/using-8-habits {#using-8-habits}

Onboarding meta-skill — explains the 8 habits, all 23 skills, and provides a decision tree for "which skill next?". Includes a complete walkthrough example (password-reset feature).

• Habit: H5 Understand First + H8 Find Your Voice
• When to use: First time using the plugin, or when unsure which skill applies
• Source: skills/using-8-habits/SKILL.md

/calibrate {#calibrate}

Self-assessment skill that determines your maturity level (Dependence → Independence → Interdependence → Significance) and writes a profile to ~/.claude/habit-profile.md. Other skills read this profile to adapt verbosity — full guidance for beginners, minimal prompts for experts.

• Habit: H8 Find Your Voice
• When to use: When plugin guidance feels too detailed or too sparse
• Deep dive: Maturity Model
• Source: skills/calibrate/SKILL.md

Compliance & Audit Skills

Governance, transparency, and regulatory readiness.

/eu-ai-act-check {#eu-ai-act-check}

EU AI Act Regulation 2024/1689 compliance checker. 9 obligations from Articles 9-15 with a tiered checklist: 25 MUST + 27 SHOULD + 8 COULD items. Includes a scope-check pre-flight to skip quickly if your system is not high-risk or not EU-targeted.

• Habit: H1 Be Proactive + H8 Spirit (Conscience)
• When to use: Before major releases of high-risk AI systems targeting the EU
• Enforcement date: 2 August 2026
• Source: skills/eu-ai-act-check/SKILL.md

/ai-dev-log {#ai-dev-log}

Generates an AI-assisted development log from git log + Co-Authored-By trailers. 4 output modes: markdown, JSON, summary, and stdout. Useful for EU AI Act Article 11 (technical documentation) audit trails and team transparency.

• Habit: H4 Win-Win + H1 Be Proactive
• When to use: Before releases, audits, or when stakeholders ask "how much was AI-generated?"
• Source: skills/ai-dev-log/SKILL.md

Quick-Select Matrix

I need to...	Use
Start a new feature from scratch	/workflow or /requirements
Understand this plugin	/using-8-habits
Adjust guidance verbosity	/calibrate
Research before deciding	/research
Review AI-generated code	/review-ai
Check all 8 habits at once	/cross-verify
Assess Body/Mind/Heart/Spirit balance	/whole-person-check
Run a focused security review	/security-check
Check EU AI Act compliance	/eu-ai-act-check
Generate AI audit trail	/ai-dev-log
Capture lessons after a task	/reflect
Deploy safely to production	/deploy-guide
Question whether the change should exist	/scrutinize
Catch drift between PRD↔design↔tasks	/consistency-check
Investigate a bug from feedback-loop	/diagnose
Write the engineering record post-fix	/post-mortem
Scaffold a project orientation hub	/save-spec
Reshape engineer content for leadership	/management-talk

Skill Anatomy

Every skill follows the same structure (source convention):

---
name: <skill-name>
description: >
  When to use this skill
user-invocable: true
argument-hint: "[arg description]"
allowed-tools: ["Read", "Glob", "Grep"]
prev-skill: <predecessor|any|none>
next-skill: <successor|any|none>
---

Body: Habit mapping → Process → Handoff → When to Skip → Definition of Done → Checkpoint.

See also

• Workflow Overview
• Habits Reference
• Architecture — how skills load and connect
• FAQ: When should I use which skill?