arc-git-pr-check

Pass

Audited by Gen Agent Trust Hub on Sep 1, 2026

Risk Level: SAFEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill's --body-file parameter in bin/run.sh provides a mechanism for reading local files and transmitting their content to GitHub via the gh pr create command. While this is an intended feature for PR management, it presents a risk if the agent is directed to read sensitive system files.
  • [DATA_EXFILTRATION]: The use of git add -A in bin/run.sh stages all modified and untracked files in the repository. If sensitive files (such as environment variables or credentials) are present and not explicitly excluded via .gitignore, they may be committed and pushed to a remote repository during the automated workflow.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides an attack surface for indirect prompt injection as it ingests and processes untrusted data from the local filesystem.
  • Ingestion points: The bin/run.sh script accepts a --body-file parameter (line 197) which reads the content of an external file to be used as the GitHub Pull Request description.
  • Boundary markers: Absent. The script does not use delimiters or warnings to prevent the agent from following instructions that might be contained within the processed PR body file.
  • Capability inventory: The skill has significant capabilities including git commit, git push, and gh pr merge (found in bin/run.sh), which can modify both the local environment and remote repositories.
  • Sanitization: There is no sanitization or validation of the content read from the --body-file before it is passed to the GitHub CLI.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 1, 2026, 05:29 PM
Security Audit — agent-trust-hub — arc-git-pr-check