daily-meeting-update
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION] (LOW): The skill relies on local CLI tools such as
gh,git, andjirato retrieve activity data. While these are standard developer tools, their invocation constitutes a command execution surface. - [DATA_EXPOSURE] (LOW): The skill accesses
~/.claude/projectsto read session history files. This involves reading local developer activity logs which, although intended for the skill's functionality, is a form of sensitive data access. - [PROMPT_INJECTION] (LOW): The skill has an indirect prompt injection surface due to its ingestion of external data from GitHub (PR titles, commits) and Jira (ticket summaries). Evidence Chain: 1. Ingestion points: Data from GitHub and Jira integrations. 2. Boundary markers: Not specified in the documentation. 3. Capability inventory: Execution of subprocesses via CLI tools. 4. Sanitization: No explicit mention of input sanitization for external strings.
Audit Metadata