skill-content-matrix
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates entirely within the local environment, reading from and writing to designated files within the agent's workspace. No remote code execution patterns, network exfiltration, or obfuscation techniques were detected.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it ingests user-supplied content pillars and local identity/style configuration files to generate its output.
- Ingestion points: User-provided pillar keywords,
identity.md,style.md,platforms.md, andaudience.mdfiles (SKILL.md). - Boundary markers: The skill does not explicitly define boundary markers for the interpolated content, relying on the Markdown table format as a structural delimiter.
- Capability inventory: The skill is limited to reading context files and writing Markdown files to the
outputs/选题矩阵/directory. - Sanitization: No specific sanitization or filtering of the input data is mentioned in the instructions.
Audit Metadata