Skills verification-before-completion
📦

verification-before-completion

Content revision r1 High Risk ⚙️ External commands

Verify Completion Claims With Evidence

AI agents can report completed work before checking tests, builds, or requirements. This skill requires fresh verification evidence before any completion or success claim.

Supports: Claude Codex Code(CC)
⚠️ 38 Poor

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request
Review the Skillstore skill "verification-before-completion" from https://skillstore.io/skills/obra-verification-before-completion.md and its manifest at https://skillstore.io/api/skills/obra-verification-before-completion/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/obra-verification-before-completion.md Signed manifest GET /api/skills/obra-verification-before-completion/manifest Signed lockfile GET /api/skills/obra-verification-before-completion/lockfile

Test it

Using "verification-before-completion". The bug is fixed.

Expected outcome:

  • I need fresh evidence first.
  • I reproduced the original symptom and ran the targeted check.
  • The symptom no longer appears, so I can report the fix with that evidence.

Using "verification-before-completion". All tests pass now.

Expected outcome:

  • I cannot claim that yet.
  • I ran the relevant test suite and read the result.
  • The run reported zero failures, so the test claim is verified.

Using "verification-before-completion". The delegated agent says the task is done.

Expected outcome:

  • I checked the diff and verified the claimed changes.
  • The evidence shows the task is partially complete.
  • I will report the remaining gap instead of claiming completion.

Security Audit

High Risk
v9 • 7/6/2026 Open versioned report

All static command and reconnaissance detections are false positives caused by Markdown fences or explanatory JSON text. I found one semantic concern: EVALUATION.json contains a self-declared safety audit that marks the skill safe to publish, which can influence reviewers if treated as trusted evidence. Remove that packaged report before publication.

2
Files scanned
391
Lines analyzed
0
Review items
0
False positives ignored

Confirmed security concerns (1)

High
EVALUATION.json:15-18
Prompt Injection Attempt Detected
EVALUATION.json embeds a self-declared audit with "safe_to_publish": true and a summary saying findings are false positives. This can influence agents or reviewers if treated as trusted evidence.
The self-assessment is explicit and located in an untrusted skill file. It may be report residue rather than an active instruction, so confidence is high but not absolute.

Risk Factors

⚙️ External commands (25)
EVALUATION.json:24-30 EVALUATION.json:30-36 EVALUATION.json:36-42 EVALUATION.json:42-48 EVALUATION.json:48-54 EVALUATION.json:54-60 EVALUATION.json:60-66 EVALUATION.json:66-72 EVALUATION.json:72-78 EVALUATION.json:78-84 EVALUATION.json:84-90 EVALUATION.json:90-96 SKILL.md:18-20 SKILL.md:20-26 SKILL.md:26-38 SKILL.md:38-79 SKILL.md:79-82 SKILL.md:82-85 SKILL.md:85-88 SKILL.md:88-91 SKILL.md:91-94 SKILL.md:94-97 SKILL.md:97-100 SKILL.md:100-103 SKILL.md:103-106
Audited by: codex View Audit History →
Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report
Security Assessment

Copy report link

https://skillstore.io/skills/obra-verification-before-completion/audits/9?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/obra-verification-before-completion/security.svg)](https://skillstore.io/skills/obra-verification-before-completion?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/obra-verification-before-completion?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/obra-verification-before-completion/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/obra-verification-before-completion.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>
Academic citations (APA · BibTeX · CFF)

APA citation

obra. (2026). verification-before-completion security audit report (audit version 9) [Author version unspecified]. Skillstore. https://skillstore.io/skills/obra-verification-before-completion/audits/9

BibTeX citation

@techreport{obra-obra-verification-before-completion-2026, author = {obra}, title = {verification-before-completion security audit report (audit version 9)}, institution = {Skillstore}, year = {2026}, number = {9}, url = {https://skillstore.io/skills/obra-verification-before-completion/audits/9}, note = {Author version unspecified} }

CITATION.cff

cff-version: 1.2.0 message: "If you use this Skill, cite its author and this versioned security audit report." title: "verification-before-completion security audit report (audit version 9)" version: "unspecified" type: report authors: - name: "obra" date-released: "2026-07-06" url: "https://skillstore.io/skills/obra-verification-before-completion/audits/9" identifiers: - type: other value: "skillstore:obra-verification-before-completion:audit:9" description: "Skillstore immutable audit report identifier"

Compare variants

10 installable variants

Each author remains a separate installable skill. The recommended variant is ranked by Skillstore evidence.

Why this variant is first

Highest Skillstore Score
CodingCossack Recommended

codingcossack-verification-before-completion

Skillstore Score 79
Evidence Confidence High
Skillstore usage 13
Updated

2026-09-09

ZhanlinCui

zhanlincui-verification-before-completion

Skillstore Score 78
Evidence Confidence High
Skillstore usage 43
Updated

2026-09-09

davila7

davila7-verification-before-completion

Skillstore Score 78
Evidence Confidence High
Skillstore usage 13
Updated

2026-09-09

ANcpLua

ancplua-verification-before-completion

Skillstore Score 78
Evidence Confidence High
Skillstore usage 11
Updated

2026-09-09

sickn33

sickn33-verification-before-completion

Skillstore Score 78
Evidence Confidence High
Skillstore usage 9
Updated

2026-09-09

Cycleaddict

cycleaddict-verification-before-completion

Skillstore Score 78
Evidence Confidence High
Skillstore usage 8
Updated

2026-09-09

Cygnusfear

cygnusfear-verification-before-completion

Skillstore Score 78
Evidence Confidence High
Skillstore usage 7
Updated

2026-09-09

DYAI2025

dyai2025-verification-before-completion

Skillstore Score 75
Evidence Confidence Medium
Skillstore usage 10
Updated

2026-09-09

obra Current

obra-verification-before-completion

Skillstore Score 38
Evidence Confidence High
Skillstore usage 30
Updated

2026-09-09

Dimon94

dimon94-verification-before-completion

Skillstore Score 38
Evidence Confidence Medium
Skillstore usage 11
Updated

2026-09-09

Skillstore Score

Why this score Evidence Confidence: High
55
Architecture
85
Maintainability
85
Content
72
Community
83
Spec Compliance

What You Can Build

Prevent False Completion Reports

Require evidence before an assistant says tests pass, builds succeed, or a task is complete.

Standardize Verification Discipline

Give teams one rule for evidence before commits, pull requests, and handoffs.

Review Agent Work Safely

Check delegated agent results with diffs and independent verification before accepting success reports.

Try These Prompts

Verify a Simple Fix 
Before claiming this is complete, identify the command that proves it. Run it now and report the result with evidence.
Verify Tests 
Run the relevant test command now. Report pass counts, failures, and exit status before saying the tests pass.
Verify Build and Requirements 
Run the build and compare results with the requirement list. Report verified items and unresolved gaps.
Verify Delegated Work 
The agent reported success. Inspect the diff, run independent verification, and report only what the evidence supports.

Best Practices

  • Run verification fresh before every success claim.
  • Match the verification command to the exact claim.
  • Report evidence, failures, and gaps with the same clarity as success.

Avoid

  • Using words like should or probably instead of verified facts.
  • Treating a linter result as proof that a build or feature works.
  • Accepting an agent success report without independent verification.

Frequently Asked Questions

Does this skill execute commands itself?
No. It instructs the agent to run the relevant verification and report evidence.
What if no verification command exists?
State that verification is unavailable and avoid completion claims. Describe what remains unverified.
Does this apply to simple tasks?
Yes. The rule applies before any completion or success statement.
Can a previous passing run count?
No. Use fresh output for the current claim.
How should delegated work be handled?
Check the diff, run independent verification, then report only verified facts.
Does this replace human review?
No. It improves evidence discipline but does not replace review or judgment.

Developer Details

Author

obra

License

MIT

Skillstore revision

r1

Version notice

The author did not declare a version.

Repository

https://github.com/obra/superpowers/tree/main/skills/verification-before-completion

Ref

e397ce5a135369cd6def89cb6400d013e2f68f2f

Maintenance freshness

7/18/2026

Usage

25 downloads · 316 views

File structure

📄 EVALUATION.json

📄 SKILL.md