completion guard procedures

Re-run tests and builds to verify requirements, quality, and state before approval.

1|1|Updated Feb 1, 2026
One-click install
npx skills add https://github.com/zb-ss/claude-plugin-workflow --skill completion-guard-procedures
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: completion guard procedures
Source: https://github.com/zb-ss/claude-plugin-workflow/tree/main/skills/phases/completion
Command: npx skills add https://github.com/zb-ss/claude-plugin-workflow --skill completion-guard-procedures

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Prevents workflow completion from proceeding when requirements are incomplete, tests/builds fail, code quality checks catch issues, or the workflow state is inconsistent.

Core Features & Use Cases

  • Independent verification: Re-runs tests and build steps rather than trusting prior gate results.
  • Code quality enforcement: Blocks completion when TODO/FIXME/HACK/XXX markers, debug statements, or secrets are present in changed files.
  • State integrity validation: Ensures all steps are marked completed and no pending objectives remain.
  • Requirements-to-code mapping: Verifies each requirement is actually implemented by checking it against the codebase.
  • Regression awareness: Re-checks the full existing test suite to avoid accidentally reverted changes.

Quick Start

Run the completion guard to independently verify requirements, quality, build success, and tests before the workflow is marked complete.

Frequently Asked Questions about completion guard procedures

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

FAQPage Schema
How do I verify workflow completion and requirements coverage before committing code?

Verify workflow completion by independently checking requirements coverage, code quality, and build/test results against the current repository state. This prevents incomplete workflows from proceeding by enforcing a strict APPROVED or REJECTED verdict.

What is the best way to prevent regressions when finishing a software development workflow?

Prevent regressions by independently rerunning the full existing test suite and build commands rather than trusting prior gate results. Validate git diffs to catch unexpected changes and ensure no accidentally reverted code reaches completion.

How do I enforce code quality checks for debug statements and secrets in changed files?

Enforce code quality by blocking workflow completion when TODO, FIXME, HACK, XXX markers, debug statements, or secrets are present in changed files. This independent check ensures final repository state integrity before approval.

Does the completion guard procedure trust previous quality gate results?

No, the completion guard procedure does not trust previous quality gate results. It independently reruns tests and build commands, revalidating prior review checks stored in state files to ensure correctness and prevent regressions.

How do I map software requirements to actual code changes for final verification?

Map requirements to actual code by independently verifying each requirement against the real codebase changes. This ensures every specified objective is genuinely implemented before the workflow is marked complete.

What happens after a workflow receives an APPROVED verdict from the completion guard?

After an APPROVED verdict, the completion guard extracts post-approval learning insights and performs cleanup actions. This finalizes the workflow state while retaining valuable verification context for future development.