Merge Risk Intelligence for Distributed Backend Systems.
Factor helps engineering teams understand what could break in production before a pull request is merged.
It analyzes pull requests in the context of your system and surfaces risks that traditional code review and CI miss.
Modern backend systems are complex, and most production issues are not caught by tests or CI.
Factor focuses on system-level risks, including:
- Hidden assumptions in code
- Failure scenarios under real conditions
- Cross-service and downstream impacts
- Missing validation and edge cases
- Business logic risks in critical workflows
Factor runs automatically on every pull request after installation.
PR opened or updated
↓
Factor analyzes code + repository context
↓
System-level risk analysis generated
↓
Results appear directly on the pull request
No manual triggers. No code changes required.
This repository contains the full documentation for Factor.
- Overview — What Factor is and why it exists
- Quickstart — Get Factor running in 5 minutes
- How Factor Works — Technical breakdown of the analysis system
- Example Reviews — Real PR analyses
- Financial Systems — Risk in monetary logic
- Authentication — Security-sensitive flows
- Background Jobs — Async system risks
- API Contracts — Integration stability
- Data Migration — Schema and state changes
- Verdicts — How to interpret risk levels
- FAQ — Common questions
- Security & Privacy — Data handling and safety
Code review and CI are good at catching:
- syntax errors
- test failures
- style issues
But they often miss:
- system-level behavior changes
- hidden assumptions
- cascading failures across services
- real-world edge cases
Factor exists to surface those risks before merge time.
To start using Factor:
- Install Factor on your GitHub account
- Select repositories to enable
- Open a pull request
Factor will automatically begin analyzing changes.
Factor is built for private source code environments.
- Only analyzes repositories you explicitly grant access to
- Does not use customer code for model training
- Does not expose code to other customers
- Follows GitHub permission boundaries
See full details in Security & Privacy docs.
With Factor, engineers don’t just review code.
They understand production risk before it ships.