github-initiative-pulse
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources such as GitHub issues and pull requests to generate reports. This constitutes a surface for indirect prompt injection if the source data contains malicious instructions aimed at the agent.\n
- Ingestion points:
SKILL.md(sync from GitHub Projects) andmodules/status-digest.md(ProjectTracker JSON payload).\n - Boundary markers: The provided markdown templates do not include explicit delimiters or instructions to ignore nested prompts.\n
- Capability inventory: The skill is designed for generating markdown digests and CSV exports; no high-privilege tools or system-level modifications are requested.\n
- Sanitization: The troubleshooting documentation mentions checking for unescaped characters in task titles, indicating basic data integrity awareness.\n- [COMMAND_EXECUTION]: The documentation instructs the use of a command-line tool,
tracker.py(e.g.,tracker.py add,tracker.py status), to manage data. This script is an external dependency and is not provided within the skill's file list.\n- [NO_CODE]: The skill is composed entirely of Markdown files providing documentation and templates; it does not ship with any executable scripts or binary files.
Audit Metadata