CtrlK
BlogDocsLog inGet started
Tessl Logo

tessl-labs/pr-review-guardrails

Evidence-first pull request review with independent critique, selective challenger review, and human handoff.

87

1.31x
Quality

92%

Does it follow best practices?

Impact

87%

1.31x

Average score across 43 eval scenarios

SecuritybySnyk

Risky

Do not use without reviewing

Overview
Quality
Evals
Security
Files

review-boundaries.mdrules/

Review Boundaries

This tile is a review aid, not an approval mechanism. Every output must make this unambiguous.

The tile must never

  • Approve a PR
  • Block a PR as final authority
  • Claim a finding is definitive without evidence
  • Pretend to know intent it was not given
  • Generate a "LGTM" or approval-equivalent signal
  • Post a summary that could be mistaken for a sign-off

Disclaimer requirement

Every tile output must include an explicit disclaimer: this is an evidence-based review aid, not an approval. The tile produces findings and questions. Humans produce decisions.

Violation test

If a developer could read the tile's output and conclude "the AI approved this, I don't need to look," the output violates this rule.

README.md

tile.json