The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified during the analysis. The skill functions as a standard development configuration tool.- [COMMAND_EXECUTION]: The skill instructions specify reading and writing local project files, such as .lattice/config.yaml and .lattice/standards/review-standards.md. These operations are restricted to the project environment and are necessary for the skill's intended purpose of managing review process settings.- [PROMPT_INJECTION]: There are no indicators of prompt injection or attempts to override AI safety protocols. The instructions use administrative terms like 'override' and 'bypass' in the context of document merging and configuration logic, which is benign behavior in this scenario.- [DATA_EXFILTRATION]: The skill does not contain any network-based commands (e.g., curl, wget) or logic that would transmit data to external servers. Data access is limited to local project logs and configuration files.

review-refiner