ai-tool-compliance

Automation skill for designing, verifying, and improving auth, cost, logging, and security compliance based on the internal AI tool mandatory implementation guide (P0/P1). Supports the full lifecycle of RBAC design, Gateway principles, Firestore policy, behavior logs, cost transparency, and the criteria verification system.

2.27x

Quality

47%

Does it follow best practices?

Impact

100%

2.27x

Average score across 3 eval scenarios

Securityby

Risky

Do not use without reviewing

Optimize this skill with Tessl

npx tessl skill review --optimize ./.agent-skills/ai-tool-compliance/SKILL.md

Evaluation results

100%

91%

Setting Up Compliance Foundation for a New AI Analytics Tool

Compliance project initialization structure

Criteria

Without context

With context

compliance-config.yaml exists

100%

rbac-matrix.yaml exists

100%

log-schema.yaml exists

100%

cost-tracking.yaml exists

20%

100%

p0-rules.yaml exists

100%

All 5 RBAC roles defined

100%

Guest has no permissions

100%

Domain weights sum to 100

20%

100%

Correct default domain weights

100%

Gate thresholds correct

100%

P0 fail override enabled

100%

Log schema required fields

30%

100%

Log schema storage tiers

16%

100%

Cost tracking AI model fields

33%

100%

33%

Implementing an AI Query Module for an Internal Analytics Tool

AI Gateway and cost logging implementation

Criteria

Without context

With context

No direct external AI API calls

100%

AI calls via server function

100%

Cache check before AI call

100%

AI call only on cache miss

100%

Model name logged

100%

Input tokens logged

50%

100%

Output tokens logged

50%

100%

Estimated cost logged

71%

100%

Reference ID returned to client

100%

No raw content in logs/responses

37%

100%

Auth check present

100%

Guest role blocked

100%

Error handler logs failures

50%

100%

44%

Auditing and Reporting Compliance for an AI Tool Codebase

Compliance verification report generation

Criteria

Without context

With context

Section 1: Summary present

100%

Section 2: Rule Results table

75%

100%

Evidence with file references

100%

Section 3: Score Breakdown

100%

Section 4: Failures Detail

62%

100%

Section 5: Gate Decision

57%

100%

Section 6: Recommendations

100%

SEC-P0-004 identified as FAIL

58%

100%

AUTH-P0-001 identified as PASS

100%

Grade is Red

25%

100%

Remediation for FAIL

25%

100%

verification-results.json produced

100%

Repository: supercent-io/skills-template
Commit: c033769

Evaluated: 28 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Setting Up Compliance Foundation for a New AI Analytics Tool Implementing an AI Query Module for an Internal Analytics Tool Auditing and Reporting Compliance for an AI Tool Codebase

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.