The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on the chub CLI tool to perform its operations. It provides instructions for the agent to execute several commands, including searching for documentation (chub search), retrieving content (chub get), saving local annotations (chub annotate), and submitting feedback (chub feedback).
[EXTERNAL_DOWNLOADS]: The chub get command is used to download documentation from an external service at runtime. This content is then used to guide the agent's responses and code generation.
[DATA_EXFILTRATION]: The skill uses the chub feedback command to transmit user-defined labels and comments to a remote service provider. While intended for product improvement, this involves sending data from the local environment to an external endpoint.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection (Category 8) because it fetches and processes untrusted documentation content without proper safeguards.
Ingestion points: Remote documentation content enters the agent's context through the output of the chub get <id> command.
Boundary markers: There are no instructions for the agent to use delimiters or ignore potentially malicious instructions embedded within the fetched documentation content.
Capability inventory: The agent can execute system commands, write data to local files (via the -o flag), store local notes, and make network requests to the feedback service.
Sanitization: The skill lacks any mechanism to validate or sanitize the documentation content before it is read and used by the agent for task completion.

get-api-docs