software-team-conflict-patterns
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: Exposure of local system information through hardcoded paths.
- Evidence: The file
references/neil-on-software-archetypes.mdcontains the absolute pathC:\Users\Big Dell\Downloads\Difficult People. This exposes the username and directory structure of the source environment. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided narratives as primary input for its analysis.
- Ingestion points: The agent accepts stories about team conflicts from the user in
SKILL.md. - Boundary markers: Absent. There are no explicit instructions to the model to ignore commands that might be embedded within the user stories.
- Capability inventory: The skill does not use tools for network access, file writing, or command execution.
- Sanitization: Absent. The skill analyzes the raw text provided by the user.
Audit Metadata