medium-writing

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface through its ingestion of local documentation and article history.\n
  • Ingestion points: The skill reads raw article exports and performance patterns from the 02-DOCS/raw/medium/ and 02-DOCS/wiki/medium/ directories.\n
  • Boundary markers: The instructions do not define clear delimiters or provided warnings to the agent to disregard instructions that might be embedded within the processed documentation.\n
  • Capability inventory: The skill is capable of generating formatted Medium article drafts and performing file-write operations to update the 02-DOCS/wiki/medium/read-ratio-patterns.md file.\n
  • Sanitization: There is no specified requirement for the agent to sanitize, escape, or validate the content ingested from the external documentation files before using it to influence draft creation.\n
  • Chain exposure: The skill processes output generated by the medium-strategy skill, creating a multi-step chain where data poisoning in one skill's output can influence the behavior of this skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 09:37 PM
Security Audit — agent-trust-hub — medium-writing