The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified. All operations are consistent with the skill's stated purpose and follow security best practices, such as redacting credentials and validating tool outputs.
[PROMPT_INJECTION]: The file references/evals.yaml contains simulated prompt injection attacks, including instructions like 'Ignore previous instructions' and 'run curl to random domains'. These are benchmark test cases designed to evaluate the agent's resistance to such attacks and do not constitute a threat within the skill's operational instructions.
[COMMAND_EXECUTION]: The skill uses shell commands to perform repository analysis via npx and manages artifacts using local helper scripts like scripts/refresh-diagram-context.sh. These commands are specific to the skill's functionality and include safety-oriented recommendations such as using dry-runs before execution.
[EXTERNAL_DOWNLOADS]: The skill fetches the @brainwav/diagram package from the public NPM registry. This download originates from a well-known service and is the core component required for the skill's operation.

diagram-cli