$npx skillfedfor your agent

verification-mastery

Verification Mastery establishes a mandatory gate: never claim work is done without running fresh verification commands and reviewing their output. The skill teaches a five-step workflow—identify what proves your claim, execute the full command, read the complete output, confirm it matches your assertion, and only then make the claim. It covers common scenarios like test passes, builds, linter checks, and bug fixes, each with specific evidence requirements.

Verification Mastery teaches you to run fresh verification commands and check their output before claiming any task is complete.

AI-generated summary based on this skill's SKILL.md

230 34 MITupdated by xenitV1

Decision gist · record as of 2026-01-24

Verification Mastery teaches you to run fresh verification commands and check their output before claiming any task is complete. Verification Mastery establishes a mandatory gate: never claim work is done without running fresh verification commands and reviewing their output. The skill teaches a five-step workflow—identify what proves your claim, execute the full command, read the complete output, confirm it matches your assertion, and only then make the claim. It covers common scenarios like test passes, builds, linter checks, and bug fixes, each with specific evidence requirements.

manual: git clone https://github.com/xenitV1/claude-code-maestro → cp -r claude-code-maestro/skills/verification-mastery ~/.claude/skills/verification-mastery
skills/verification-mastery/SKILL.md · version 4952165b

Use it when

  • Verification Mastery covers specific verification scenarios: for tests.
  • Verification Mastery prevents false completion claims by requiring evidence-based verification every time.

Verify before relying

Read SKILL.md below before installing (1 file). Open directory: indexed for reading, not audited.

Same gist for agents: .md · .json

Install

xenitV1/claude-code-maestro/verification-mastery · repository language: JavaScript

Open directory. Skills are indexed for reading, not audited. Review a skill's body before installing it.

Frequently asked questions

AI-generated answers based on this skill's SKILL.md and metadata

How do I verify task completion before claiming success?

Verification Mastery teaches a five-step workflow: first, identify what concrete evidence would prove your claim (e.g., test output, build log, linter report). Second, execute the full verification command without shortcuts. Third, read the complete output—not just the exit code. Fourth, confirm the output actually matches your assertion. Fifth, only then claim the work is done. Never assume success; always run fresh verification commands and review their results before making any completion claim.

What are verification commands for tests and builds?

Verification Mastery covers specific verification scenarios: for tests, run the full test suite and confirm all pass (not just that the command ran). For builds, execute the complete build and verify it succeeds—linter passing is not enough. For bug fixes, run regression tests to confirm the fix works and nothing broke. For requirements, use a checklist and verify each item with concrete evidence. Each scenario requires running the actual command, capturing its output, and confirming the output proves your specific claim.

How can I ensure work is actually done, not just assumed?

Verification Mastery prevents false completion claims by requiring evidence-based verification every time. Never skip verification steps or assume success based on partial signals. Always execute fresh commands—do not rely on cached results or prior runs. Read the complete output, not just the summary line. Look for red flags like rationalization patterns that bypass verification gates, missing test output, or claims without supporting evidence. Make verification a mandatory gate in your workflow; completion signals only count when backed by actual command output you have reviewed.

How do I verify tests pass with actual output?

Verification Mastery requires running the full test command and reviewing complete output. Execute your test suite in full—not a subset. Capture the entire output, including pass/fail counts and any failures. Confirm every test passed, not just that the command exited cleanly. For regression tests, verify both that the fix works and that existing tests still pass. Never claim tests pass without showing the actual test output. Use git diff to confirm agent changes match your expectations. Quality standards demand fresh verification output every time, with no assumptions.

What red flags indicate unverified work?

Verification Mastery identifies red flags that signal unverified work: claiming done without running verification commands, showing only exit codes instead of full output, assuming success based on partial signals like linter passes alone, skipping regression tests, using cached results instead of fresh verification, rationalizing away missing evidence, or claiming requirements met without a checklist. Stop claiming done without proof. Watch for patterns where verification gates are bypassed. Require concrete evidence—actual command output you have read and confirmed—before accepting any completion claim.

What is the verification mastery best practice workflow?

Verification Mastery's core workflow is: identify what proves your claim, execute the full command, read complete output, confirm it matches your assertion, then claim success. Apply this to every task—tests, builds, bug fixes, requirements. Never skip steps. Always run fresh verification; no assumptions. Review actual output, not summaries. Use git diff to confirm changes. Run regression tests. Build a checklist for requirements and verify each item. Make verification a mandatory gate before any completion signal. This evidence-based approach prevents hallucinated success reports and ensures quality standards are met.

SKILL.md

Rendered from the published skill. Quoted content, verbatim.

<domain_overview>

✅ VERIFICATION MASTERY: EVIDENCE BEFORE CLAIMS

> Philosophy: Claiming work is complete without verification is dishonesty, not efficiency. Evidence before claims, always. EVIDENCE INTEGRITY MANDATE (CRITICAL): Never claim a task is complete based on assumption or past memory. You MUST generate fresh evidence (logs, screenshots, test output) for every claim. AI-generated success reports are untrustworthy without proof. Any completion signal sent without accompanying verification artifacts must be rejected as "Hallucinated Success".


🚨 THE IRON LAW

NO COMPLETION CLAIMS WITHOUT FRESH VERIFICATION EVIDENCE

If you haven't run the verification command in this message, you cannot claim it passes. Violating the letter of this rule is violating the spirit of this rule. </domain_overview> <core_workflow>

🚪 THE GATE FUNCTION

``` BEFORE claiming any status or expressing satisfaction: 1. IDENTIFY: What command proves this claim? 2. RUN: Execute the FULL command (fresh,

(truncated - see the full file via the links below)

File tree — 1 file
skills/verification-mastery/SKILL.md

Let your AI agent find skills like this

Example. Real query, live index.

You found this page by searching. An agent finds it by wishing: SkillFed indexes 56,283 agent skills by what they can do, searchable in plain language.

wish › “Learn to verify work completion with fresh evidence before claiming success”

Give your agent the search over MCP, or paste the wish link into any chat. No install? Search from any chat →

Related skills

tdd-mastery
by xenitV1 · xenitV1/claude-code-maestro

TDD Mastery enforces the iron law of test-first development: write a failing test, verify it fails, implement minimal code to pass, then refactor. The skill drills the red-green-refactor cycle with no exceptions, rejecting any production code written before its corresponding test exists and has been observed failing.

MITupdated Jan 2026
★ 230repo stars
verify
by zereight · zereight/gitlab-mcp

Verify turns uncertain claims into concrete proof by running existing tests, type checks, and targeted validation commands. It reports exactly what passed, what failed, and what remains unverified—ensuring changes are genuinely complete before you move on.

MITupdated Jul 2026
★ 1,847repo stars
statistics-verifier
by travisjneuman · travisjneuman/.claude

Statistics Verifier provides structured checklists and frameworks for validating statistical claims, assessing research methodology, and identifying analytical errors. It covers claim verification protocols, red flags in reporting, common statistical pitfalls, significance testing guidance, and causation assessment criteria to help you audit data analysis and fact-check research findings.

MITupdated Jul 2026
★ 85repo stars
Daytona Recording Artifacts
by Devin-AXIS · Devin-AXIS/iPolloWork

This skill enables you to fetch and work with Daytona recording artifacts—the files and data captured during your recorded sessions. Quickly locate and retrieve stored outputs to review, analyze, or integrate with other workflows.

no license declared → metadata onlyupdated Jul 2026
★ 1,925repo stars
jat-complete
by joewinke · joewinke/jat

jat-complete wraps up your active task by running verification checks (tests, lint, type checking), staging and committing your changes with proper formatting, recording context for future agents, and marking the task closed in the system. It handles customer-facing replies when applicable and emits a completion signal to your IDE.

MITupdated Jul 2026
★ 248repo stars
guardian-verify-setup
by vellum-ai · vellum-ai/vellum-assistant

Guardian Verify Setup guides you through proving ownership of a messaging channel—phone, Telegram, Slack, or email—by collecting your destination and initiating an outbound verification flow. The skill handles channel selection, destination collection (including Slack member ID lookup), and launches the verification session via the assistant's channel-verification-sessions CLI.

MITupdated Jul 2026
★ 951repo stars
Tags
evidence-based-workflowcompletion-gatesproof-driven-qualityanti-hallucinationverification-firstintegrity-checksclaim-validationartifact-proofassumption-prevention