sync-agenticat

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt's hard rule "Report any secret read during the harvest, no matter why" directs the agent to disclose secrets encountered, which can force the LLM to output secret values verbatim.

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

• Third-party content exposure detected (high risk: 0.85). The required workflow step “Resolve the source” (local path / owner/repo → github URL / git URL → clone / raw URL → fetch) and “Harvest” then reads outsider-authored free text from those fetched/cloned repositories (e.g., their SKILL.md, agents/*.md, commands) into the agent’s LLM context for transformation.