How does the Council Gate handle model disagreements?

It calculates a confidence score based on model consensus; if the score falls below your defined threshold, the skill returns an 'UNCLEAR' status (Exit Code 2), allowing the pipeline to pause for human intervention.

Where are the AI deliberation logs stored?

All deliberations, including model responses and final synthesis, are saved locally in the .council/logs/ directory, which can be uploaded as CI/CD artifacts for auditing.

Can I focus the quality check on specific areas like security?

Yes, the skill supports a rubric-focus parameter that allows you to target specific domains such as Security, Performance, or general Maintainability during the audit.

Is this skill compatible with GitHub Actions?

Absolutely. It is designed for native integration with GitHub Actions, and the documentation provides a ready-to-use YAML configuration for automated workflows.

Council Quality Gate

Name: Council Quality Gate
Author: amiable-dev

byamiable-dev

•

Deployment & DevOps

Automates CI/CD pipeline approvals and quality checks using multi-model AI consensus to enforce rigorous code standards.

The Council Quality Gate skill integrates the LLM Council's multi-model consensus mechanism directly into your CI/CD workflows, providing an intelligent decision layer for pull requests and deployments. By evaluating code snapshots against specific rubrics such as security or performance, it generates high-confidence verdicts that can automatically pass pipelines, fail them, or flag them for human review. This skill ensures that complex architectural and security decisions are validated by multiple AI models, providing a detailed audit trail of every deliberation for complete transparency in automated DevOps processes.

Key Features

01Multi-model consensus for high-confidence automated approvals

02Native GitHub Actions support with configurable confidence thresholds

03Deterministic exit codes for seamless CI/CD integration

04Detailed deliberation transcripts and audit logs for every check

057 GitHub stars

06Customizable quality rubrics targeting security or performance

Use Cases

01Gating production deployments on security-focused AI audits

02Automating Pull Request approvals based on multi-model verification

03Enforcing consistent quality standards across large DevOps environments

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add amiable-dev/llm-council bundled

For use in Claude.ai and ChatGPT

Download Skill