github
Fail
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: HIGHCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
- [CREDENTIALS_UNSAFE]: A hardcoded fallback API key "BETA_TEST_KEY_MARCH_2026" is present in the SKILL.md file and across all tool scripts (e.g., scripts/get_me.py, scripts/add_issue_comment.py, scripts/create_repository.py) as a default value for the router authentication.
- [PROMPT_INJECTION]: The skill presents a significant attack surface for indirect prompt injection. 1. Ingestion points: Untrusted content is fetched from external repositories and issues using tools like issue_read (scripts/issue_read.py), get_file_contents (scripts/get_file_contents.py), and search_code (scripts/search_code.py). 2. Boundary markers: There are no delimiters or instructions to ignore embedded commands within the fetched data. 3. Capability inventory: The skill has high-impact write permissions, including the ability to create or update files (scripts/create_or_update_file.py), push multiple files (scripts/push_files.py), and merge pull requests (scripts/merge_pull_request.py). 4. Sanitization: Content retrieved from GitHub is not sanitized or validated before being processed by the agent.
Recommendations
- AI detected serious security threats
Audit Metadata