daily-meeting-update
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes data from external and local sources which may contain untrusted content.
- Ingestion points: Data is collected from GitHub (commits, pull requests, and reviews), Jira (assigned tickets), and local Claude Code session history via the
claude_digest.pyscript reading from~/.claude/projects. - Boundary markers: The skill's instructions do not define specific delimiters or instructions to prevent the agent from following directions potentially embedded in commit messages or ticket titles.
- Capability inventory: The skill can execute local shell commands and a Python script, and it generates text that influences the subsequent interview process.
- Sanitization: There is no evidence of sanitization or validation of the text retrieved from external APIs or local session logs.
- [COMMAND_EXECUTION]: The skill uses several local CLI tools and a provided Python script to gather context.
- Evidence: The skill executes
gh auth status,jira,git, andpython3 ~/.claude/skills/daily-meeting-update/scripts/claude_digest.py. These executions are documented as part of the integration detection and data gathering phases and are preceded by user consent prompts.
Audit Metadata