write-for-agents
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of instructional markdown and metadata. No executable scripts, command-line operations, or network requests were detected.\n- [INDIRECT_PROMPT_INJECTION]: The instructions direct the agent to read external documents and referenced files as part of its process. This creates an ingestion surface for data that the agent does not control.\n
- Ingestion points: Step 1 in
SKILL.mdrequires reading documents, pointers, and files.\n - Boundary markers: None explicitly defined.\n
- Capability inventory: None defined in the skill itself.\n
- Sanitization: No sanitization steps are specified.\n- [NO_CODE]: The skill does not include any source code, scripts, or binary files.
Audit Metadata