puukis
UserA personal Claude Code environment. Persistent memory, loop detection, 15 skills, and lifecycle hooks. macOS, Linux, and Windows.
Categories
Indexed Skills (28)
changelog
Generate a CHANGELOG.md entry from git log since last tag — invoked via /changelog
debrief
End-of-session structured reflection — what worked, what broke, what to do differently. Run before /exit.
orchestrate
Intelligent task routing — evaluates complexity and asks the user whether to handle in the main session or dispatch to specialist sub-agents. Entry point for any medium or large task. Run this before /build or /engineer on anything non-trivial.
remember
Store a user-confirmed structured learning or durable observation in lstack memory.
receipt
Manage Change Receipts from Claude Code — wraps lstack brain receipt CLI
work
Start structured AI coding work with Brain overview, Repo Passport, Context Governor, Change Receipts, and AI Mistake Firewall awareness
analytics
Show lstack memory analytics, including structured learning health.
architect
System design before any code is written — invoked via /architect; produces ADRs and ARCHITECTURE.md
blueprint
Generates a .blueprint.md spec file before any code is written. Agents implement against the blueprint, not vague prompts.
build
Plan then implement with an explicit approval gate between phases — invoked via /build
careful
Enable opt-in warnings or denials for risky Bash commands in the current Claude Code session.
debug
Systematic root cause analysis — reproduce, isolate, hypothesize, verify before any fix
docs
Writes accurate documentation by reading source first — never invents API signatures
engineer
Writes production-ready code — reads patterns first, runs tests after, no gold-plating
forget
Delete structured learnings or legacy observations from lstack memory.
freeze
Restrict Edit, Write, and MultiEdit to approved paths for the current Claude Code session.
guard
Enable maximum safety workflow by combining careful safety mode with a freeze edit boundary.
interrogate
Socratic requirement clarification — one question at a time before any planning or code. Auto-activates when a request is vague, large, or unclear.
learn
Manually store a typed, trust-aware structured learning in lstack.
parallel
Spawn isolated sub-agents across git worktrees for genuinely independent tasks
planner
Structured task breakdown with clarifying questions before planning — no code written
recall
Search structured learnings and legacy observations in lstack memory.
refactor
Improves code structure without changing behavior — tests before and after each step
reviewer
Code review producing only numbered actionable findings — no rewrites, no praise
security
Security audit — scans for secrets, injection, bad deps, auth issues; reports findings only
ship
Pre-ship checklist that gates the release — runs 5 checks and blocks until all pass
passport
Show or refresh Repo Passport and context — wraps lstack brain passport/context/overview CLI
unfreeze
Clear the current Claude Code session freeze boundary.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.