ai-slop-gate/ai-slop-gate
Contribute to ai-slop-gate — an open-source CI/CD policy gate that detects AI-generated code slop, security issues, and compliance violations. Use when adding a new LLM provider (Gemini, Groq, Ollama) or static provider, writing analysis rules, extending the policy engine, working with reporters, fixing compliance sidecar logic, or writing tests. Activate when working with ai_slop_gate/ source tree, policy.yml, snapshot, or GitHub Actions workflows.
97
97%
Does it follow best practices?
Impact
100%
1.19xAverage score across 3 eval scenarios
Advisory
Suggest reviewing before use
Evaluation results
100%
24%Add Mistral LLM Provider to ai-slop-gate
New LLM provider structure and JSON contract
Correct file location
100%
100%
Extends LlmProvider
100%
100%
kind = 'llm'
100%
100%
analyze() implemented
100%
100%
analyze_pr() implemented
100%
100%
Prompt file location
100%
100%
Prompt enforces top-level 'issues' key
0%
100%
Prompt prohibits markdown/extra text
100%
100%
Prompt enforces snake_case signal
0%
100%
Prompt enforces severity enum
75%
100%
Registry registration shown
100%
100%
CLI option shown
75%
100%
100%
8%Add a Dead Link Detector Static Provider and Review Cache Behaviour
Cache layer design and static provider conventions
Correct static file location
100%
100%
Extends BaseProvider
100%
100%
kind = 'static'
100%
100%
collect() implemented
100%
100%
ProviderObservation fields correct
100%
100%
Test positive case
100%
100%
Test negative case
100%
100%
Cache LLM-only rule
100%
100%
Cache key components
53%
100%
Engineer proposal rejected
100%
100%
No provider-specific logic in cache
100%
100%
Duplicate call prevention
75%
100%
100%
14%Refactor ai-slop-gate to Separate Compliance Checking from Analysis
Architecture principles and compliance sidecar placement
Engine IO violation identified
100%
100%
CLI thinness violation identified
100%
100%
Compliance sidecar placement identified
100%
100%
Engine is pure in refactored example
100%
100%
Compliance runs after providers
100%
100%
Compliance runs before reporters
100%
100%
Compliance profiles mentioned
0%
100%
Compliance activation flags
28%
100%
Correct exit code for blocking
100%
100%
Compliance checks described
100%
100%
CLI only wires components
100%
100%
No compliance inside engine
100%
100%
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6