markdown-to-html
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external markdown content which could contain malicious instructions.
- Ingestion points: markdown source material processed by the skill as described in skill.md.
- Boundary markers: none explicitly defined in the high-level workflow instructions.
- Capability inventory: file system inspection and network tool calling via sandbase_run.
- Sanitization: skill.md explicitly mandates that the agent must sanitize untrusted html.
- [DATA_EXFILTRATION]: the skill interacts with the sandbase api using vendor-specific tools such as sandbase_run. this involves transmitting data to an external service. the risk is mitigated by explicit instructions to minimize data transfer and require user confirmation before any data transmission or paid calls.
Audit Metadata