Awesome Reviewers

Verification · Intent and spec conformance

Output guardrails

Validates what an automated reviewer or fixer produces before it is posted or committed: schema, policy and safety checks.

Sub-techniques

  • Schema validation V1058.001

    Reject malformed structured output and retry.

  • Policy rails V1058.002

    Block outputs that leak secrets, change protected files or exceed scope.

How to build it

  1. Define the output schema and validate every response.
  2. List protected paths and actions the bot may not take.
  3. Log every blocked output for review.

Signals in a repository

  • Guardrail config files or validators around model calls.