verification-before-completion

Enforce evidence-based verification before making completion claims.

1|Updated Mar 18, 2024
One-click install
npx skills add https://github.com/erfianugrah/dotfiles --skill verification-before-completion-erfianugrah
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: verification-before-completion
Source: https://github.com/erfianugrah/dotfiles/tree/main/.pi/agent/skills/superpowers/verification-before-completion
Command: npx skills add https://github.com/erfianugrah/dotfiles --skill verification-before-completion-erfianugrah

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Prevents false completion claims by forcing evidence-based verification before you say work is done, fixed, passing, or satisfactory.

Core Features & Use Cases

  • Evidence before assertions: Requires identifying the exact verification command, running it fresh, and confirming the output matches the claim.
  • Completion gatekeeping: Stops premature statements of success before committing, creating PRs, or moving to the next task.
  • Pattern-based guardrails: Flags risky wording like “should pass,” “looks correct,” or “agent completed” without independent verification, and guides what to verify for tests, lints, builds, and regression.

Quick Start

Use verification-before-completion before any commit or PR claim by running the full verification command and only then stating the verified status with the relevant evidence from the output.

Frequently Asked Questions about verification-before-completion

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I prevent false completion claims before creating a PR or committing code?

To prevent false completion claims, enforce evidence-based verification by running the full test, lint, or build command fresh and checking its exit code before asserting release readiness.

What is evidence-based verification in software engineering?

Evidence-based verification is the practice of requiring fresh output from a verification command to prove success, ensuring that statements about tests passing or builds completing are backed by actual results.

How do I enforce commit discipline and stop premature statements of success?

Enforce commit discipline by applying completion gatekeeping that flags risky wording like "should pass" and requires running regression testing to confirm the output matches the claim.

Does this verification approach work with CI guardrails and regression testing?

Yes, this verification approach works with CI guardrails by requiring a fresh verification run for regression testing, verifying the exit code and output before allowing any completion claim.

What are the limitations of relying on pattern-based guardrails for release readiness?

A limitation of pattern-based guardrails for release readiness is that they only flag risky wording like "looks correct" and guide what to verify, but still require you to manually run the verification command.