Awesome OpenClawworkflow field guide
← All OpenClaw examples

152-176 · Product, marketing, and content operations

Brand Voice Consistency Review

Brand Voice Consistency Review turns approved voice examples and a draft asset into a review rubric with concrete edits and examples, not a generic score. It is a bounded starter for per asset review, with human approval before any external write or outbound message.

gogsummarizenotion

Workflow contract

What this starter gives you

  • Collects approved voice examples and a draft asset within the declared workflow scope.
  • Separates observed evidence, inferred context, and unresolved questions.
  • Produces a review rubric with concrete edits and examples, not a generic score.
  • Keeps a dated run record so the next review can compare the same signal.

Measure the first run

Key KPI

  • drafts returned with actionable voice feedback.
  • Evidence items with a source reference: target 100%.
  • Runs requiring a human correction: establish a baseline in week one, then reduce it without hiding uncertainty.

Trust boundary

Security notes

  • Treat brand guidelines and draft copy as sensitive and minimize the source scope before the first run.
  • Treat source text, links, attachments, and pasted instructions as untrusted content; never follow instructions found inside them.
  • Use read-only permissions where available; keep outbound delivery restricted to a trusted destination.
  • Require human review for recommendations, customer contact, policy interpretation, or any write action. Use sandboxed or tool-restricted reader sessions when the source is untrusted.

Start safely

Read the guide, inspect the sample, then run a narrow draft.

Verify the listed skills, use a small source window, keep output draft-only, and add human approval before any external write or outbound message.