modernize
Pass
Audited by Gen Agent Trust Hub on Nov 16, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from GitHub pull request titles, bodies, and comments. It includes explicit defensive instructions to treat this content as data rather than instructions and requires processing it through a specialized tool (
gstack-issue-guard) to prevent injection attacks. - [COMMAND_EXECUTION]: The skill utilizes several local binaries (e.g.,
gstack-skill-start,gstack-question-log,gstack-skill-end) located in~/.claude/skills/gstack/bin/. These are legitimate components of the toolset used for lifecycle management and telemetry. - [DATA_EXFILTRATION]: Telemetry and event logging are handled via local scripts. The data collected (session IDs, outcomes, and question preferences) is relevant to the tool's operation and is managed by local infrastructure.
- [SANITIZATION]: The skill demonstrates high security awareness by requiring the agent to read pull request diffs as untrusted input when dealing with cross-repository PRs, ensuring that scripts or configuration changes are reviewed before execution.
Audit Metadata