antislop-writing

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill directs the agent to build an "exact-strings list" and copy verbatim numbers, names, quoted phrases, code identifiers, and links from the user's draft into the inventory/output, which would force the LLM to reproduce any secrets (API keys, tokens, passwords) present in the input.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The skill’s intake/composition workflow operates on the user-supplied draft text (and its explicit inventory of copied verbatim strings/links), so outsider-authored free text is directly ingested at runtime.