Architecture-aware pull request intelligence
Catchrisky codebefore it ships.
> analyzing PR #42... risk: HIGH — confidence: 94%|
GrepAI builds repository context before review, traces architecture impact normal code review misses, and posts merge risk directly inside GitHub without changing developer workflow.
No install required. Connect a repository and review risk inside GitHub.
< 60s
Analysis per PR
Median review latency from diff to risk signal
0
Workflow changes
Works inside the PR flow your team already uses
100%
GitHub native
Risk comments, context, and merge guidance stay in-thread
AI
Powered analysis
Repository context graph built before judgment lands
Review artifact
Impact trace
Findings
- Shared authentication middleware now shifts session validation across dependent services
- Token refresh boundary mutation touches gateway request state and billing execution paths
- Circular auth provider dependency increases bootstrap risk before downstream retries
Recommendation
Review architecture boundaries before merge. Validate retry flows and downstream session integrity.
Files changed
- src/auth/auth.middleware.ts
- src/auth/token-refresh.ts
- src/session/session-validator.ts
- src/gateway/request-context.ts
- src/billing/billing.service.ts
Blast radius
How GrepAI works
01
Event signal
PR Event
PR event streamed from GitHub.
02
Context engine
Diff + Context
Repository context graph generated.
03
Risk model
AI Analysis
Impact and boundary risk evaluated.
04
PR feedback
GitHub Feedback
Risk review posted inside the PR.