Agents edit sensitive files
AI coding agents can open PRs and edit production-sensitive files.
AI PR Protection for GitHub
Stop AI-generated PRs reaching production without proof.
Inntris adds a required GitHub Action check that PASS/BLOCKs AI-generated pull requests and creates a verification receipt for every decision.
The problem
CI can pass while the action still violates production policy.
Build checks are not intent checks
CI can pass even when the agent was not allowed to touch that surface.
Reviewers need proof
Teams need evidence of what the agent attempted and why it was allowed or blocked.
The demo
AI PR -> Inntris required check -> PASS/BLOCK receipt
AI PR
An AI coding agent opens a pull request against the production branch.
Inntris required check
The GitHub Action reads the changed files, policy, and optional Promptfoo risk evidence.
PASS/BLOCK receipt
Reviewers get a decision and verification evidence showing what happened and why.
What gets protected
Start with the files buyers are nervous about.
auth
payments
secrets
database migrations
infrastructure
GitHub workflows
production branch
Starter
$200
per month per repo
- 1 repo
- 1 protected branch
- GitHub Action setup
- basic policy file
- PASS/BLOCK receipts
- monthly review summary
- setup included for first 3 design partners
Trust-safe wording
Adds an AI-specific policy gate. Creates verification evidence. Supports review and audit.
Keep normal CI and code review.
Use it as evidence for review, audit, and production change control.