linkedin-content-planner
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to ingest external data from 'Voice samples' (past posts) to calibrate the writing style, creating a surface where malicious instructions could potentially influence behavior.\n
- Ingestion points: The
SKILL.mdfile defines 'Voice samples (optional): paths to past posts' as a valid input for voice calibration.\n - Boundary markers: The instructions lack explicit delimiters or 'ignore' directives to separate the content of the voice samples from the primary task instructions.\n
- Capability inventory: The skill is limited to text generation (markdown and optional JSON). It does not have access to tools for file system modification, network communication, or shell command execution.\n
- Sanitization: There are no provided instructions or mechanisms to sanitize, validate, or escape the content of the ingested voice samples before the agent processes them.
Audit Metadata