Tooluck logoTooluck
SKILL.md

verify

Agent skill fromcodeaholicguy/ai-devkit

AI DevKit · Enforce evidence-based completion claims — require fresh command output before reporting success. Use when completing any task, fixing a bug, finishing a phase, running tests, building, deploying, or making any "it works" claim.

How to install verify

install verify
# install the ai-devkit pack with the skills CLI
npx skills add codeaholicguy/ai-devkit

# or add only this skill to Claude Code: copy the folder

skills/verify/

# into

~/.claude/skills/verify/ (all projects)

.claude/skills/verify/ (one project)

skills/verify/SKILL.md

Apache-2.0Full file
1# Verify
2
3Prove it works before saying it works.
4
5## Hard Rules
6
7- Do not claim completion without fresh terminal evidence from this session.
8- Forbidden words in completion claims: "should", "probably", "seems to", "likely", "I believe", "I think it works". These signal unverified assertions.
9- Cached, remembered, or previous-session output is not evidence. Run it again.
10- A change that alters anything a user or caller sees requires a `dev-smoke` run, whatever the claim is called.
11
12## Gate Function
13
14Every completion claim must pass all steps in order:
15
160. **Classify** — Does the change alter anything a user or caller sees (CLI output, API response, UI, displayed status or state, file formats)? If yes, `dev-smoke` is required evidence alongside tests.
171. **Identify** — What command proves this claim? If multiple commands are needed, run the gate once per command.
182. **Run** — Execute the full command now. No partial runs, no skipping.
193. **Read** — Read complete output. Check exit code. Count pass/fail.
204. **Confirm** — Does the output prove the exact claim?
215. **Report** — State the result in the Report Format below.
22
23If any step fails, stop. Fix the issue and restart from step 0.
24
25If no verification command exists (e.g., no test suite), tell the user and ask them how to verify before claiming done.
26
27## Verification Patterns
28
29| Claim | Required Evidence | Not Sufficient |
30|---|---|---|
31| Tests pass | Test output: 0 failures, exit 0 | Previous run, "should pass now" |
32…

Excerpt shown under the repository's Apache-2.0 license.

More skills in ai-devkit

Keep exploring

Similar skill repositories

thedotmack /

claude-mem
116k

Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More

CCCXGHGM+135 skills

Newsletter

Join the Community

Subscribe to our newsletter for the latest news and updates