The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on the pica CLI tool to perform all operations, including listing connections, searching for API actions, and executing them on external platforms.
Evidence: Commands such as pica connection list, pica actions search, and pica actions execute are used throughout SKILL.md to interact with the Pica infrastructure.
[PROMPT_INJECTION]: The skill presents a potential surface for indirect prompt injection as it instructs the agent to ingest documentation ('knowledge') from the CLI and use it to construct subsequent API execution commands. However, this is the intended functional workflow and the instructions include advice on using single quotes for JSON values to prevent shell-related issues.
Ingestion points: Output from pica actions search and pica actions knowledge (SKILL.md).
Boundary markers: Absent.
Capability inventory: pica actions execute allows the agent to perform network requests via the Pica proxy.
Sanitization: The instructions recommend using single quotes to avoid shell escaping issues when passing JSON data.

pica-actions