jake-speak
Pass
Audited by Gen Agent Trust Hub on Jul 27, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill is designed to create and save a visual HTML recap file to the user's desktop (defaulting to
~/Desktop/). This requires the agent to perform file system write operations outside the project repository. - [DATA_EXFILTRATION]: The skill includes security-conscious instructions to ensure generated HTML files remain self-contained. It explicitly forbids external network requests and requires that all assets, including CSS, JavaScript, and images, be inlined or embedded as data URIs. This prevents the recap files from leaking data to external servers.
- [INDIRECT_PROMPT_INJECTION]: The skill processes technical data from the repository, such as commit messages and code changes, to generate its summaries. There is a minor risk that malicious content within the repository could influence the summary's narrative, although the skill's focus on high-level outcomes and avoidance of "code mechanics" provides a natural layer of abstraction.
- [DYNAMIC_EXECUTION]: The skill generates an HTML document at runtime by populating a local template (
assets/recap-template.html) with summarized content. This is a standard document generation task and does not involve the execution of untrusted code.
Audit Metadata