Features
Entity-level diff analysis highlighting structurally risky changes
Graph-centric risk scoring based on dependents and blast radius
Change classification (text, syntax, functional) using ConGra taxonomy
Local execution with no LLM or API calls required for core triage
Optional LLM review via `inspect review` (Anthropic, OpenAI, Ollama, any OpenAI-compatible server)
Supports 21 programming languages plus 5 data formats via sem-core and tree-sitter
MCP server with 6 tools for AI-agent integration (inspect_triage, inspect_entity, etc.)
Parallel graph building for large codebase support (e.g., Sentry: 40s → 4s)
Supports Git diff, PR review, single file review, and repo-wide benchmark
Multiple output formats: terminal, JSON, markdown
No database required; stateless per commit
95% recall on planted bugs benchmark (141 bugs, 52 PRs, 5 repos)
83.5% high/critical recall on Greptile benchmark
Four-phase pipeline: Extract, Classify, Score, Group
Untangles tangled commits via Union-Find on dependency edges
Autonomous vulnerability scanning and PR generation
PR review for quality, security, and best practices
Legacy COBOL to Java migration with testing
Incident triage and root cause analysis
Agent Canvas: manage multiple agents in one workspace
Cloud agents run even when your machine is off
Schedule recurring tasks via cron or event triggers
Trigger workflows from GitHub, Slack, PagerDuty
Model-agnostic: supports Claude Code, Codex, Gemini CLI, MiniMax M2.7
Isolated sandbox execution on VM or cloud
Fine-grained access control and budget guardrails
Audit trails for enterprise compliance
Self-hosted deployment in your VPC
Open-source SDK for custom agent development
In-chat model switching with saved LLM profiles
