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
OS-level automatic memory capture of code, docs, chats, meetings
LTM-2 Long-Term Memory Engine storing 9 months of workflow history
Natural language search across all memories
1-click save and AI-tagging of code snippets
Workstream Activity timeline view in desktop app
MCP integration for context with GitHub Copilot, Cursor, Claude, Goose
Real-Time Context for favorite LLMs (cloud or local)
Chrome plugin for capturing research links and highlights
VS Code, JetBrains, Visual Studio, Sublime Text plugins
JupyterLab, Obsidian plugins
Meeting capture and automatic standup report generation
Private by design: local by default, air-gapped, cloud optional
End-to-end control: on-device processing, no external servers
Time-based memory queries for precise recall
