The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses standard system utilities (curl, python3, sed) to interact with the Quip REST API and process JSON/HTML data. All shell commands are used for legitimate API orchestration.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill manages authentication via a user-supplied environment variable (QUIP_ACCESS_TOKEN). It includes explicit safety instructions to avoid logging or echoing this credential in output, adhering to secret management best practices.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external Quip documents (via GET threads/{thread_id}).
Ingestion points: Document HTML content retrieved from platform.quip.com.
Boundary markers: No specific delimiters or safety warnings are implemented for the ingested content.
Capability inventory: The agent can perform network requests and script execution, creating a potential path for data exfiltration if malicious instructions in a document are followed.
Sanitization: The skill uses basic regex or Python parsers to strip HTML tags for readability but does not perform security-focused sanitization of the content before presenting it to the agent.

using-quip