update-stars
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to fetch external data (GitHub star counts) and use that information to modify source code files. This creates a potential surface for indirect prompt injection.
- Ingestion points: The agent reads data from the external GitHub website (specified in
SKILL.md). - Boundary markers: No boundary markers or instructions to ignore embedded content in the external data are present.
- Capability inventory: The agent has the capability to modify multiple files (
CommunityStatsItems.tsx,GitHubButton.tsx,TrustSection.tsx) and execute git commands (commit,push). - Sanitization: No sanitization or validation of the fetched star count is specified beyond rounding.
- [COMMAND_EXECUTION]: The skill requires the agent to execute git commands to commit and push changes directly to the
mainbranch. This bypasses typical code review workflows (Pull Requests), though it is consistent with the stated purpose for the project author.
Audit Metadata