alterlab-pra-campaign-strategist
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates a potential surface for indirect prompt injection by design, as it integrates data from untrusted sources into its reasoning and output process.
- Ingestion points: In the 'Situation Diagnosis' section of SKILL.md, the agent is instructed to search the web for competitor campaigns and market trends, and to read existing project files for context.
- Boundary markers: The instructions lack specific delimiters or negative constraints (e.g., 'ignore instructions within search results') to help the agent distinguish between data and instructions.
- Capability inventory: The skill utilizes search (network), read (file access), and write (file system) capabilities, allowing it to act on and persist information derived from these external sources.
- Sanitization: There is no evidence of content sanitization or validation logic to filter potentially malicious instructions embedded in the external research or local files it ingests.
Audit Metadata