Skip to content
Back to skills

Verify Done

ASecurity

Gate completion claims with observed build, test, runtime, documentation, and diff evidence. Use at the end of implementation or when the user asks whether existing work is complete. Do not use as the primary code-review workflow, and do not repair failures during a verification-only request unless the user separately authorizes implementation.

  • 2 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added September 19, 2026
ai-agentscode-reviewdocumentation

Security analysis

A100/100

Scanned September 19, 2026

npx -y skills add adityaarakeri/senior-agent-skills --skill verify-done --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Verify Done?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Verify Done
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/adityaarakeri-verify-done/badge)](https://www.skillsdirectory.com/skills/adityaarakeri-verify-done)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: verify-done
description: Gate completion claims with observed build, test, runtime, documentation, and diff evidence. Use at the end of implementation or when the user asks whether existing work is complete. Do not use as the primary code-review workflow, and do not repair failures during a verification-only request unless the user separately authorizes implementation.
---

# Verify Done

Claim only what the available evidence demonstrates.

## Select the Mode

- **Implementation completion:** Run the gate, repair in-scope failures while implementation authority remains active, then rerun the failed and neighboring checks.
- **Verification only:** Run safe relevant checks, report failures or blocked gates, and stop without changing implementation or configuration.

## Evidence Gate

1. Inspect the repository's actual command definitions, worktree baseline, required services, cost, credentials, side effects, and expected runtime.
2. Build, compile, bundle, or import the changed code where applicable.
3. Run focused tests, the relevant suite, and any lint or type checks. Report exact pass, fail, and skip counts and distinguish pre-existing failures.
4. Exercise the assembled user-facing behavior on a safe realistic input. Prefer a dry run, sandbox, fixture, or disposable environment. Do not mutate production, incur charges, send messages, expose private data, or expand authority merely to satisfy this gate.
5. Audit status and the complete diff for unintended files, debug code, generated artifacts, secrets, and unrelated churn.
6. Confirm documentation, configuration, migrations, rollout, rollback, and monitoring remain consistent with changed behavior.
7. Record checks that could not run because of credentials, hardware, services, platform, or environment. Never turn an unavailable check into a pass.

## Report

State the command or action, observed result, and residual gap. If a gate fails during verification-only mode, report the task as not verified and stop. If repair is in scope, fix the cause and repeat the gate before claiming completion.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.