skills.rest
zardusai-cyber
zardusai-cyberCOMMUNITY
Explore all skills by this author →
GitHub

skill-comply

Generates compliance specs and evaluates agents' tool-call behavior with scored reports.

1|1|Updated Mar 31, 2026
Search Tags:#reporting#claude-code#compliance-testing#tool-calls#scenario-generation#skill-compliance#temporal-order
One-click install
npx skills add https://github.com/zardusai-cyber/zardus_setup --skill skill-comply-zardusai-cyber
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: skill-comply
Source: https://github.com/zardusai-cyber/zardus_setup/tree/main/ecc/skills/skill-comply
Command: npx skills add https://github.com/zardusai-cyber/zardus_setup --skill skill-comply-zardusai-cyber

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires pyyaml>=6.0, and includes scripts (resource) components.

What problem does it solve?

Automates the verification of coding agents against predefined skills, rules, and agent definitions by generating observable specs, running multi-level prompts, executing agent sessions, and producing a structured compliance report.

Core Features & Use Cases

  • Auto-generates a formal compliance spec from a skill document.
  • Generates three scenario prompts with decreasing prompt strictness and runs agent sessions.
  • Classifies tool calls against the expected workflow and reports deterministic compliance scores with timelines.
  • Recommends hook promotions when compliance falls below the threshold.

Quick Start

Run the skill-comply pipeline to generate a spec, execute three scenario tests, and produce a final compliance report.

Frequently Asked Questions about skill-comply

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

FAQPage Schema
How do I automate compliance testing for my coding agent's tool calls?▼

Agent compliance testing works by generating scenario prompts with decreasing strictness, running agent sessions, and classifying tool calls against an expected workflow. It measures adherence through deterministic temporal checks and reports a timeline-rich compliance score.

How do I verify my agent follows a specific workflow and temporal order during execution?▼

You verify workflow adherence by running agent sessions against generated scenario prompts and classifying tool calls. The process measures temporal order through deterministic checks and outputs a timeline-rich compliance score indicating how well the agent followed the spec.

What can I do when my coding agent's compliance score falls below an acceptable threshold?▼

When compliance falls below the threshold, the testing pipeline recommends hook promotions to enforce rules. This ensures the agent adheres to the required workflow and improves future scenario test results through structural enforcement rather than prompt reliance.

Does this compliance testing pipeline require any specific dependencies or environments to run?▼

The pipeline requires the pyyaml dependency, version 6.0 or higher, to function. It operates by generating a compliance spec from a skill document and executing scenario tests through agent sessions to produce the final structured report.

What is the best way to generate a compliance spec from an existing skill document?▼

The best way to generate a compliance spec is to run the skill-comply pipeline, which auto-generates a formal spec from your skill file. It then uses this spec to execute three scenario tests and produce a structured compliance report.

Related Skills

Explore More Software Engineering Skills→
MubasherMohammedMubasherMohammed

skill-builder

Generate ClawBio skill directories with SKILL.md, Python skeleton, and tests from JSON or YAML spec files.

Community
Intermediate
web-infra-devweb-infra-dev
12.8k

rspack-sftrace

Trace Rust function calls and convert them to perfetto protobuf traces.

Official
Advanced
tenlisboatenlisboa
7

product-management

Define governance workflows that scope tasks, write specs, and delegate coding.

Community
Advanced
samymitysamymity
1

skill-comply

Grade coding agent tool-call traces against expected sequences for skill compliance.

Community
Advanced