Verify software changes through tests, typecheck, and manual validation.
npx skills add https://github.com/kamelmh/opencode --skill verify-kamelmh
Please help me install this Agent Skill. Skill: verify Source: https://github.com/kamelmh/opencode/tree/main/skills/verify Command: npx skills add https://github.com/kamelmh/opencode --skill verify-kamelmh
This Skill prevents false confidence by turning vague claims like “it should work” into concrete, checkable evidence.
Use the verify skill to validate a proposed fix by asking it to identify the exact behavior that must be proven and then run the narrowest available tests or commands, reporting clear pass/fail results.
High-intent search queries and answers about installing and using this skill.
To verify a bug fix, you must provide concrete evidence that the change works by running existing tests, performing typecheck/build, executing narrow command checks, and finally applying manual validation with explicit pass/fail reporting.
Evidence-first verification for refactoring turns vague claims like "it should work" into checkable proof by prioritizing existing tests, typecheck/build, narrow command checks, and manual validation to demonstrate correctness before completion.
You prove a proposed software change works through transparent reporting that summarizes what passed, what failed, and what remains unverified, relying on an evidence-first workflow instead of unvalidated claims.
Yes, this verification approach applies to feature work, bug fixes, and refactors where correctness must be demonstrated before completion, applying an evidence-first workflow that prioritizes existing tests and typecheck/build.
To verify a refactor, follow a safety-prioritized verification order: start with existing tests, then typecheck/build, then narrow command checks, and finally manual validation when automation is insufficient.
You report CI confidence safely by summarizing what passed, what failed, and what remains unverified without bluffing, ensuring transparent evidence reporting instead of relying on unvalidated claims for completion.
sswapnil2Run verification commands and evaluate outputs before claiming completion.
mkrlabsGenerate structured QA verdicts and metrics for validation workflows.
h-4vokRun lint, typecheck, test, smoke, and build checks across a repository and emit a pass/fail report.
wecode-aiRun shell commands and manipulate files inside an isolated E2B sandbox.
