repos-branch-policy-audit
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingest data from external Azure DevOps API endpoints, including project names, repository names, and policy details. If these resources are named maliciously by an attacker with access to the Azure DevOps environment, they could theoretically influence the agent's behavior during processing.
- Ingestion points: API responses from
/_apis/projects,/_apis/git/repositories, and/_apis/git/policy/configurations(in SKILL.md). - Boundary markers: The skill does not define specific delimiters for separating API-provided strings from its own instructions.
- Capability inventory: Shell command execution via
curl(restricted to GET requests) and markdown table generation. - Sanitization: The skill includes instructions for URL encoding project names and special characters in shell commands to prevent command injection.
Audit Metadata