The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes npx to download and execute the official CLI tool from the vendor, @pachca/cli. This is a documented and standard method for using the platform's developer utilities.
[COMMAND_EXECUTION]: Multiple instructions and reference files describe the execution of the pachca command-line utility to perform actions like opening forms, sending messages, and creating tasks. These operations are essential to the skill's primary function of bot automation.
[INDIRECT_PROMPT_INJECTION]: The skill processes data originating from external webhooks and interactive form submissions. This ingestion of potentially untrusted user content represents an attack surface.
Ingestion points: External data enters the agent context through the data object in view_submission events received via webhooks.
Boundary markers: The skill provides instructions on mapping data keys to block names, which helps in maintaining structured data handling.
Capability inventory: The agent can perform actions such as message creation and task management using the pachca CLI based on the data received.
Sanitization: The documentation includes limits for metadata and title lengths, and specifically provides a mechanism (HTTP 400 or submit-response --errors) for performing field validation and returning errors.

pachca-forms