team-manager
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill interacts with external data sources such as GitHub API responses (team names, member lists, PR activity) and a workspace configuration file (
.github/agents/preferences.md), creating a surface where untrusted data could influence the agent's behavior. - Ingestion points: Data retrieved via GitHub REST/GraphQL API tools and the
.github/agents/preferences.mdfile. - Boundary markers: Absent; the instructions do not specify delimiters or warnings to ignore instructions embedded in the ingested data.
- Capability inventory: The skill utilizes GitHub API tools to modify organization membership, invite users, and write records/reports to the local filesystem.
- Sanitization: No explicit sanitization or validation of input data is mentioned, although the behavioral rules requiring explicit user confirmation for all changes serve as a manual mitigation.
Audit Metadata