verification-before-completion
Validate before marking done
Install verification-before-completion
npx skills add obra/superpowers --skill verification-before-completion Skills install into ~/.claude/skills/ and load on demand. Other SKILL.md tools (Cursor, Codex, Gemini CLI) use the same files.
What it does
Forces the agent to verify a fix actually works before declaring the task complete — fewer “done” claims that aren’t. obra/superpowers debugging skill.
- bug fixes
- refactors
Trust & safety
Score 90/100 · High trust- From the widely-used obra/superpowers collection.
- Manually reviewed and verified by Loadout.
- No network access — runs locally only.
- Markdown-only instructions, no bundled scripts.
Heuristic signal based on provenance and capabilities — not a security audit. Always review a skill’s SKILL.md and scripts before first run.
Scan any SKILL.md yourself →Works with
- Claude Code
- Claude.ai
- Claude API
- Cursor
- Codex CLI
- Gemini CLI
Related skills
FAQ — verification-before-completion skill
What does the verification-before-completion skill do?
Forces the agent to verify a fix actually works before declaring the task complete — fewer “done” claims that aren’t. obra/superpowers debugging skill.
How do I install the verification-before-completion skill?
Run: npx skills add obra/superpowers --skill verification-before-completion. Or git clone the repo into ~/.claude/skills/. Skills load automatically when relevant — no extra config.
Which agents support verification-before-completion?
It reads the SKILL.md standard, so it works with Claude Code, Claude.ai, Claude API, Cursor, Codex CLI, Gemini CLI.
Is the verification-before-completion skill safe to run?
High trust (trust score 90/100). From the widely-used obra/superpowers collection. Manually reviewed and verified by Loadout. No network access — runs locally only. Markdown-only instructions, no bundled scripts.