The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The script scripts/npm_client.py hardcodes the API token storage path to /root/.npm-token.json. This design assumes the agent is running with root privileges and attempts to write to a restricted system directory, violating the principle of least privilege.
[DATA_EXPOSURE]: API authentication tokens are stored in a predictable, plain-text JSON file (/root/.npm-token.json). Predictable credential locations facilitate unauthorized access if the system or a parallel process is compromised.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data from the NPM REST API (proxy host domain names, forward hosts, and configurations) and displays it to the agent. This content could contain malicious instructions designed to influence the agent's behavior.
Ingestion points: API responses from nginx/proxy-hosts and nginx/certificates in scripts/npm_client.py.
Boundary markers: Absent; the data is printed directly to the console for the agent to read.
Capability inventory: File system write access (/root/.npm-token.json), Network operations (urllib.request).
Sanitization: No evidence of sanitization or validation of strings retrieved from the API before they are printed.

npm-proxy