The Agent Skills Directory

[COMMAND_EXECUTION]: The skill performs HTTP POST requests to crm.psquared.dev and notifications.psquared.dev. These endpoints are used to query CRM data and trigger configuration checks on demo agents. This behavior is consistent with the skill's stated purpose and targets the developer's own infrastructure.
[DATA_EXPOSURE]: The skill reads the local .env file to retrieve authentication tokens (EMAIL_DRAFT_ONLY_BEARER and PSQUARED_CRM_TOKEN). Accessing environment files for specific service tokens is a standard practice for local agent integrations and does not indicate malicious intent in this context.
[INDIRECT_PROMPT_INJECTION]: The skill processes structured data (JSON and GraphQL responses) from external APIs, which represents a potential injection surface.
Ingestion points: API responses from crm.psquared.dev and notifications.psquared.dev containing company names and status messages (SKILL.md).
Boundary markers: Absent; the skill does not wrap the external content in specific delimiters.
Capability inventory: File system read access (to .env) and outbound network requests (SKILL.md).
Sanitization: No explicit sanitization or validation of the API data is mentioned before it is reported to the user.

sanity-check