kiro-verify-completion

Validates task, fix, test, and feature completion claims against fresh command evidence.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/scapia-oss/compass --skill kiro-verify-completion-scapia-oss
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: kiro-verify-completion
Source: https://github.com/scapia-oss/compass/tree/main/tools/cc-sdd/templates/agents/claude-code-skills/skills/kiro-verify-completion
Command: npx skills add https://github.com/scapia-oss/compass --skill kiro-verify-completion-scapia-oss

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? AI coding agents often claim work is done based on stale test runs, subagent reports, or partial checks. This Skill blocks false completion claims by requiring fresh, scope-matched evidence before any task, fix, test, or feature is declared complete. ## Core Features & Use Cases - Claim-Type Gates: Verifies four claim types — TASK, FIX, TEST_OR_BUILD, and FEATURE_GO — each with its own evidence requirements. - Fresh Evidence Enforcement: Requires actual command output and exit codes from the current code state, rejecting stale or inferred results. - Structured Verdicts: Returns VERIFIED, NOT_VERIFIED, or MANUAL_VERIFY_REQUIRED with the claim, evidence, and any scope gaps. - Use Case: Before reporting a feature as ready for GO, run this Skill to confirm the full test suite passed, the built artifact boots to a usable state, requirements are covered, and no blocked tasks remain. ## Quick Start Ask the agent to verify the claim that the notification feature is complete using the FEATURE_GO claim type before marking it done.

Frequently Asked Questions about kiro-verify-completion

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

FAQPage Schema
How do I verify a task is complete before marking it done?

Provide the exact claim, its type (TASK, FIX, TEST_OR_BUILD, or FEATURE_GO), and fresh validation command output with exit codes. The Skill returns VERIFIED only when the evidence matches the claim's scope, otherwise NOT_VERIFIED or MANUAL_VERIFY_REQUIRED.

What evidence is required for a FEATURE_GO claim?

FEATURE_GO requires the full test suite result, a runtime smoke boot showing the built artifact reaches its first usable state, requirements coverage, cross-task integration, design alignment, and blocked-task assessments. A passing test suite alone is not sufficient.

Can I trust a subagent's success report as verification?

No. Reported success from another subagent is not verification evidence. The Skill requires fresh command output and exit codes from the current code state before accepting any completion claim.

When does verification return MANUAL_VERIFY_REQUIRED?

It returns MANUAL_VERIFY_REQUIRED when no canonical validation command is known, the required environment is unavailable, or a mandatory manual verification step cannot be executed by the agent.

Why does a passing build not prove a feature works?

Build and lint success do not prove runtime behavior. Type erasure, module loading, native ABI, and boot-time configuration issues can still fail at runtime, which is why FEATURE_GO also requires a runtime smoke boot result.