电商宣传整改清单

Use this skill after claim-risk-audit to turn detected risk flags into a practical handoff for a merchant, editor, or compliance owner.

Prerequisites

• - A claimriskaudit.json generated from redacted facts.
• The owner\'s current product facts, approved claims, and target-platform profile.

Input and output

Input is an audit JSON. The output carries only claim references, detected risk flags, rewrite instructions, evidence tasks, and manual sign-off fields; it does not copy customer text.

Output is claimremediationkit.json with task priority, safe rewrite direction, evidence requirements, and fixed externaleditexecuted=false.

Procedure

1. 1. Run claim-risk-audit and preserve the audit as the evidence record.
2. Run python scripts/buildclaimremediation.py --input claimriskaudit.json --output claimremediationkit.json.
3. A human owner edits the original content in an approved editor, checks current platform rules, and signs off before any external action.

Error recovery

Malformed input or missing humanreviewrequired fields exits non-zero. The planner never invents a replacement claim, guarantee, price, or legal conclusion.

Permission and privacy boundaries

This skill does not edit or upload content, publish, remove a review, hide a comment, approve a claim, or certify compliance. It does not access accounts or customer data.

Acceptance

The example must map absolute wording, guarantee, evidence, brand, contact, and privacy flags to concrete tasks, and every item must remain MANUALSIGNOFF_REQUIRED.