title: Integrations description: Connect Cadence to Git platforms, AI providers, AI assistants, and automation workflows order: 3

Integrations

Cadence provides four integration surfaces: a webhook server for continuous Git platform monitoring, AI providers (OpenAI and Anthropic) for enhanced detection, an AI skills system for embedding detection into agents and pipelines, and a plugin API for extending detection with custom strategies.

Integration Surfaces

AI Providers

Connect Cadence to OpenAI or Anthropic to add AI-powered analysis on top of pattern-based detection. AI analysis runs after suspicious items are flagged, providing confidence scores, reasoning, and detailed indicators.

Best for:

Deeper analysis of flagged commits with natural-language explanations
Higher confidence scoring using LLM reasoning
Report summarization and pattern explanation

Guides: Overview · Configuration · Examples

Webhook Server

Run Cadence as a long-running HTTP server that receives push events from GitHub or GitLab, queues analysis jobs, and returns results via REST or Server-Sent Events.

Best for:

Continuous repository monitoring on every push
CI/CD pipeline integration
Self-hosted Git platforms (GitHub, GitLab, Gitea, Forgejo)

Guides: Overview · GitHub Setup · GitLab Setup · API Reference

Agent Skills

Cadence exposes four internal AI skills and two external skills (via skills.json) compatible with Claude, ChatGPT, and custom automation tools.

Best for:

AI coding assistants invoking detection directly in conversation
On-demand analysis without a running server
Embedding detection results into automation pipelines

Guides: Overview · Internal Skills · External Skills

Analysis Plugins

Extend Cadence's detection pipeline by implementing the StrategyPlugin interface. Plugins register with the PluginManager and participate in all analysis runs.

Best for:

Custom detection strategies unique to your codebase or policy
Domain-specific heuristics not covered by built-in detectors
Wrapping third-party analysis tools in the Cadence pipeline

Guides: Overview · Writing Plugins

Quick Start

Start the Webhook Server

Bash

cadence webhook --port 8000 --secret "your-webhook-secret"

Enable AI Analysis

Bash

export CADENCE_AI_ENABLED=true
export CADENCE_AI_PROVIDER=openai   # or: anthropic
export CADENCE_AI_KEY=sk-...
cadence analyze /path/to/repo -o report.json --config .cadence.yaml

Run On-Demand Analysis via API

Bash

curl -X POST http://localhost:8000/api/analyze/repository \
  -H "Content-Type: application/json" \
  -d '{"url": "https://github.com/owner/repo", "branch": "main"}'

Documentation