Skills Reference

Skills Reference

8-habit-ai-dev ships 24 user-facing markdown skills. Skills are read-only guidance: they structure investigation, planning, review, deployment, communication, and reflection, but they do not execute changes by themselves.

Tip

If you are unsure where to start, use /using-8-habits or the quick selector below.

Quick Selector

Situation	Start with
New feature	/requirements
Unclear problem	/research
Architecture decision	/design
AI-generated diff	/review-ai
Security-sensitive change	/security-check
Production deploy	/deploy-guide
Operational finding	/operational-state
Bug with unclear cause	/diagnose
Incident/config hotfix drift	/consistency-check
After a validated fix	/post-mortem
Need stakeholder-ready wording	/management-talk

Workflow Skills

/research {#research}

Investigate before specifying. Produces a research brief with findings, constraints, open questions, and recommended next step.

• Page: Step 0 · Research
• Habit: H5 Seek First to Understand
• Source: skills/research/SKILL.md

/requirements {#requirements}

Define what, why, who, scope, success criteria, and Definition of Done before implementation.

• Page: Step 1 · Requirements
• Habit: H2 Begin with End in Mind
• Source: skills/requirements/SKILL.md

/design {#design}

Surface architecture options and make the human decision explicit.

• Page: Step 2 · Design
• Habit: H8 Find Your Voice
• Source: skills/design/SKILL.md

/breakdown {#breakdown}

Decompose approved work into atomic, ordered tasks.

• Page: Step 3 · Breakdown
• Habit: H3 Put First Things First
• Source: skills/breakdown/SKILL.md

/build-brief {#build-brief}

Read existing code and produce an implementation brief before editing.

• Page: Step 4 · Build Brief
• Habit: H5 Seek First to Understand
• Source: skills/build-brief/SKILL.md

/review-ai {#review-ai}

Review AI-generated changes before commit, with findings ordered by severity.

• Page: Step 5 · Review AI
• Habit: H4 Think Win-Win
• Source: skills/review-ai/SKILL.md

/deploy-guide {#deploy-guide}

Plan staging, production, rollback, verification, and reconciliation for deploys.

• Page: Step 6 · Deploy Guide
• Habit: H1 Be Proactive
• Current note: includes CI/CD proof-layer discipline and provider-managed canary/capacity reconciliation gates.
• Source: skills/deploy-guide/SKILL.md

/monitor-setup {#monitor-setup}

Add or verify health checks, alerting, metrics, logging, dashboards, and runbooks after deployment.

• Page: Step 7 · Monitor Setup
• Habit: H7 Sharpen the Saw
• Source: skills/monitor-setup/SKILL.md

Assessment And Review Skills

/cross-verify {#cross-verify}

Run a 17-question 8-Habit readiness check across a plan, diff, or implementation.

• Use before implementation commitment, PR creation, or risky merge.
• Source: skills/cross-verify/SKILL.md

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

Assess Body, Mind, Heart, and Spirit balance for a feature, component, or PR.

• Use when a technically correct change may still be unbalanced.
• Source: skills/whole-person-check/SKILL.md

/security-check {#security-check}

Run a focused security lens across secrets, input handling, auth, dependencies, config, infrastructure, and source-of-truth drift.

• Use for code, infrastructure, alerting, email, webhook, container, environment, and mounted-config changes with risk-bearing behavior.
• Source: skills/security-check/SKILL.md

/scrutinize {#scrutinize}

Question whether the proposed change should exist, then trace the actual path to verify it does what it claims.

• Use before committing to a plan or before merging a PR.
• Source: skills/scrutinize/SKILL.md

/consistency-check {#consistency-check}

Check drift across PRD, design, and tasks. Also supports incident/config hotfix mode for symptom, evidence, root cause, fix, deploy path, and verification consistency.

• Use after persisted spec artifacts or before closing an operational hotfix PR.
• Source: skills/consistency-check/SKILL.md

/reflect {#reflect}

Capture a short post-task lesson, including what helped, what confused, and which skill was missed.

• Use after meaningful work, especially after incidents or release work.
• Source: skills/reflect/SKILL.md

Operations And Debugging Skills

/operational-state {#operational-state}

Classify operational findings as Watch, Fix Candidate, Active Incident, Resolved, Handoff, Known Accepted Issue, False Positive, or Self-Resolved before action.

• Use before mutating production state, closing an alert/issue, accepting a known problem, or handing off ownership.
• Boundary: read-only guidance; no runtime state engine or automatic production change.
• Source: skills/operational-state/SKILL.md

/diagnose {#diagnose}

Investigate a bug through feedback loop, reproduction, hypothesis, instrumentation, fix with regression test, and cleanup.

• Use when the cause is not obvious.
• Source: skills/diagnose/SKILL.md

/post-mortem {#post-mortem}

Write the engineering record after a validated fix: root cause, mechanism, fix, validation, and how it slipped through.

• Use after a reliable repro, known cause, identified fix, and validated outcome exist.
• Source: skills/post-mortem/SKILL.md

Onboarding, Memory, And Communication Skills

/workflow {#workflow}

Walk through the 7-step workflow interactively.

• Use when starting a feature and you want the agent to prompt step by step.
• Source: skills/workflow/SKILL.md

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

Explain the plugin and route user intent to the most relevant skill.

• Use for onboarding or when a request does not map clearly to one skill.
• Source: skills/using-8-habits/SKILL.md

/calibrate {#calibrate}

Assess guidance maturity and write a local verbosity profile for Claude Code.

• Use when guidance feels too verbose or too sparse.
• Source: skills/calibrate/SKILL.md

/save-spec {#save-spec}

Scaffold a project-root SPEC.md orientation hub when a repository needs a durable project digest.

• Use when starting in an unfamiliar repo and the repo fits the project-orientation hub mode.
• Source: skills/save-spec/SKILL.md

/management-talk {#management-talk}

Reshape engineer-to-engineer content for leadership channels such as JIRA, Slack, standup, email, or meeting notes.

• Use after technical investigation when the audience changes.
• Source: skills/management-talk/SKILL.md

Audit And Governance-Adjacent Skills

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

Redirect stub for EU AI Act compliance work. The canonical implementation lives in pitimon/claude-governance.

• Use the governance plugin for the full checklist.
• Boundary: this plugin does not certify compliance.
• Source: skills/eu-ai-act-check/SKILL.md

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

Generate an AI-assisted development log from git history and co-author trailers.

• Use for transparency, release review, or audit trail preparation.
• Source: skills/ai-dev-log/SKILL.md

See Also

• Workflow Overview
• Maturity Model
• Architecture