product-marketing
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it reads external repository content (README, landing pages, package.json) to generate summaries.
- Ingestion points: Scans local repository files including
README.md,package.json, and various marketing assets. - Boundary markers: The instructions do not define specific delimiters for separating the ingested file content from the agent's own prompt instructions.
- Capability inventory: The skill is restricted to local file system read operations and writing a markdown file to the
.agents/directory. - Sanitization: The workflow includes a mandatory manual review and correction step where the user must approve the drafted content before it is saved, providing a human-in-the-loop check against malicious content.
- [SAFE]: The skill uses standard project management practices by storing context in a dedicated hidden directory (
.agents/). No network operations, credential access, or dynamic code execution patterns were found.
Audit Metadata