Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Get Started Free →

A5c AI/Verification — measured AI agent skill | DecimalAI

a5c-ai/verification

SkillScore18Maintained

Verification-before-completion discipline ensuring all success criteria are met, tests pass, and reviews complete before declaring work done.

by a5c-ai

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

QA & Test Engineering pack

decimalai skills pull a5c-ai-verification

Get this skill — curl · raw file

curl -fsSL https://app.decimal.ai/s/a5c-ai-verification/SKILL.md

Raw file: /s/a5c-ai-verification/SKILL.md ↗ · machine-readable summary at /s/a5c-ai-verification.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 9d ago

—-45% tokens0% turns18% of whole cases pass with skill (4/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-03	✗→✓	▲ Improved	-19%	0%
case-18	✗→✓	▲ Improved	-67%	0%
case-20	✗→✓	▲ Improved	-65%	0%
case-21	✗→✓	▲ Improved	-54%	0%
case-01	✗→✗	＝ Same ✗	-54%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 7/31/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

a5c-ai / babysitter

library/methodologies/rpikit/skills/verification/SKILL.md

View on GitHub ↗

License MITImported 7/31/2026Version v1

Skill contents

SKILL.md

At phase checkpoints
Before final completion declaration
When validating that all acceptance criteria are met

Verification Checklist

All automated tests pass
Manual verification steps confirmed
Code review completed (APPROVE or APPROVE_WITH_NITS)
Security review passed
Plan document updated with completion status
All success criteria from the plan are satisfied
No untracked deviations from the plan

Key Rules

Never declare done without running verification
Document any verification failures
Verification failures halt progress until resolved

Tool Use

Integrated into methodologies/rpikit/rpikit-implement (step execution and completion)

Want to see if a5c-ai/verification would activate on your phrasing? Try it before you install — no LLM call, no signup.

Other measured skills in the registry, with their headline benchmark lift.