Guard Verification
Use when code adds a guard. Prove tests catch its loss.
Install
npx skills add SamueleBassoli/hermes-fleet-apps --skill guard-verificationgit clone --depth 1 https://github.com/SamueleBassoli/hermes-fleet-apps.git /tmp/hermes-fleet-apps
mkdir -p ~/.claude/skills
cp -r /tmp/hermes-fleet-apps/roles/qa/skills/guard-verification ~/.claude/skills/guard-verificationPersonal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.
From the SKILL.md
# Guard Verification A green suite proves the code passes the tests. It does **not** prove the tests would catch the code being wrong. When the deliverable is guarded behaviour — a refusal, a safety check, a branch that changes where data goes — the only evidence that a guard is protected is: delete it, and watch the test that owns it fail. Use this whenever a change adds a rule the system must obey, a check that refuses something, or a branch that changes an outcome. Skip it for purely additive code with no guard in it — there is nothing whose absence would be silent. ## Before building a refusal: measure its blast radius A guard that refuses an input class (skip an input with 2+ subjects, reject a file over N) can throw away inputs that work today. Before writing it, run a throwaway read-only script over the REAL data the guard will see and print two numbers: - **how many inputs it would refuse**: if that is near zero, the simple rule is safe. If it is large, the rule needs a sharper condition, for example counting only items above the existing quality floor so a blurred bystander does not trigger it. - **how many of those were actually WRONG under the old behaviour**: check against ground truth that is already there (the same entity's other, unambiguous inputs). This turns "it might be wrong" into "it was wrong in 2 of 3". The same pass also finds bad records that already exist, and those need to be reported to the owner separately.
Details
- Skill name
guard-verification- Repository
- SamueleBassoli/hermes-fleet-apps
- Path
roles/qa/skills/guard-verification/SKILL.md- License
- MIT
- Repo updated
- 31 October 2026
FAQ
What does the Guard Verification skill do?
Use when code adds a guard. Prove tests catch its loss.
How do I install the Guard Verification skill?
skills CLI (Claude Code, Cursor, Codex and more): npx skills add SamueleBassoli/hermes-fleet-apps --skill guard-verification. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.
Is Guard Verification popular?
It lives in SamueleBassoli/hermes-fleet-apps, which has 44 GitHub stars and 0 forks.
Related skills
Agent Facing CLI
44SamueleBassoli/hermes-fleet-apps
Use when an agent will drive a CLI. Contract test first.
Agent Fleet Design
44SamueleBassoli/hermes-fleet-apps
Use when adding, backing up, or driving the agent fleet.
Agent Identity Setup
44SamueleBassoli/hermes-fleet-apps
Use when setting up or resetting agent SOUL, memory, skills.
QA Test and Fix
430xranx/agentbrief
When the user wants to find and fix bugs, or says 'QA this,' 'test this,' 'find bugs,' 'why is this broken,' 'it doesn't work,' 'check for bugs,' 'smoke test,' or after any significant code change. This is the full QA cycle: discover → reproduce → diagnose → fix → verify.
Fleet Project Pipeline
44SamueleBassoli/hermes-fleet-apps
Use when running a product through the fleet: stages, gates, kanban cards, briefs, legs, qa fix pairs, merges and the clean-up sweep.
Regression Testing
430xranx/agentbrief
When the user wants to prevent regressions, improve test coverage, or says 'add tests,' 'improve coverage,' 'we keep breaking this,' 'write regression tests,' 'characterization tests,' or after fixing a production bug to ensure it never recurs.
Browse the most popular Claude skills, all testing & qa skills, or write your own with the skills generator.