The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the bun runtime to execute local TypeScript tools including PangramScore.ts, StatSignals.ts, and WatermarkScan.ts located in the ~/.claude/LIFEOS/TOOLS/ directory. It also performs local network requests via curl to localhost:31337 for notification purposes. These operations are restricted to the local environment and the skill's defined tooling.
[EXTERNAL_DOWNLOADS]: The skill communicates with the external Pangram API at pangram.com to retrieve empirical AI detection scores. User-supplied text is sent to this service for analysis as part of the core functionality.
[PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection because it ingests and processes untrusted user-supplied text. 1. Ingestion points: Text is provided by the user via command-line arguments or stdin to the analysis tools. 2. Boundary markers: The skill does not define specific delimiters to isolate the text being analyzed from agent instructions. 3. Capability inventory: The skill has access to subprocess execution and local network communication. 4. Sanitization: The skill documentation does not mention any sanitization or filtering of the content being analyzed.

DetectAI