research
Pass
Audited by Gen Agent Trust Hub on Jul 8, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exposes a surface for indirect prompt injection (Category 8) by processing untrusted data from external sources.\n
- Ingestion points: The skill specifically instructs the agent to investigate "primary sources" such as official documentation, source code, and specs from external URLs (
SKILL.md).\n - Boundary markers: There are no defined delimiters or instructions to ignore embedded commands within the researched content to prevent the agent from obeying malicious instructions found in documentation or source code.\n
- Capability inventory: The agent has the capability to execute
ghCLI commands to list/create labels and create GitHub issues using the researched data (SKILL.md).\n - Sanitization: While the skill uses a "body-file" for issue creation (which mitigates some command injection risks), it lacks logic to sanitize or filter the content of the research findings before they are processed by the agent or used to generate summaries.
Audit Metadata