daily-summary
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a shell script located at a static path on the local filesystem.
- Evidence: The instructions explicitly require running
~/.claude/scripts/daily-summary.shto gather activity data. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the local development environment, including PR titles, issue comments, and commit messages, which are then used to generate a markdown summary.
- Ingestion points: Data is read from the output of the shell script (redirected to
/tmp/daily-summary.json) and from project configuration inCLAUDE.md. - Boundary markers: The skill does not define specific delimiters or warnings to prevent the agent from following instructions potentially embedded in the fetched activity data.
- Capability inventory: The skill has the ability to execute shell scripts and write files to the
~/code/claude-log/directory. - Sanitization: There is no evidence of sanitization or validation performed on the PR titles or commit messages before they are processed by the agent.
Audit Metadata