All 94 skillsQuality3 of 13
stryx:verify
When Claude is about to say done, fixed or ready to commit: fresh proof comes first.
/stryx:verifyInside its piece: 11 specks, one for each file on its shelf. Its twelve group-mates glow on the open book behind it. The light splits toward orange, the band of Quality.
In numbers
counted 2026-10-12 in the installed set- 9reference files
- 11files in the skill
- 129KB on disk
- 11chapters
- 1agent works by this book
Use it when
in the skill’s own words- when about to say work is done
- fixed
- passing
- complete
- ready or safe to commit
- before a commit
- push or PR
- before ticking a task or saying "100% done"
- before moving to the next task
- delegating
- handing back to the user
- When the user says
- verify
- check
- check if it works
- validate changes
- fresh check
- run verification
- run QA
- quality gate
- handoff
- finished or complete
- Also use when setting up Stop or SubagentStop completion hooks
Not for
another skill does it- Finding why a check fails→ stryx:debug
What it covers
7 parts- the fresh-evidence Iron Law
- which evidence each claim needs
- done criteria per task type (feature, bug fix, refactor, UI, performance, a document, PDF or printed design)
- the ten quality gates
- fresh-context verification through stryx:verify-agent
- auto-fix limits
- evidence-first reports
Inside the book
11 chapters, by title only- The Iron Law
- The gate - five steps, every time
- Which evidence each claim needs
- Regression test: the red-green proof
- Choose the depth
- Step 1 in practice - find the right commands
- Rules for every task type
- Auto-fix limits
- Evidence files (Windows-safe, never /tmp)
- Reporting - evidence first
- Override - only on the user's words
Its helper
1 agent works by this book- stryx:verify-agentWhen a "done" claim needs checking by someone who did not write the code.
Works with
the skills it names most- stryx:debug→Something is broken: a bug, an error, a crash, a failing test or wrong data.
- stryx:test→When tests must be written or fixed, or you want the test written first (TDD).
- stryx:ai-power→When you are not sure which skill fits, or you say "load aipower": it is the map of all skills.
- stryx:design-audit→When checking a design that exists: a UX walk-through, accessibility, polish or screen sizes.
- stryx:eval→When you must measure how a prompt, an agent or an AI feature behaves.
- stryx:execute→When a written plan must be carried out to the end, or you say "keep going until it passes".
Claude loads it by itself when the moment comes; by hand, type the line at the top. It comes with the whole set: Installs all 94 skills, 32 agents and the brain program as one set. The skill’s own text is private source and is not shown here.