The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the clay CLI tool to perform operations like listing workbooks, querying tables, and fetching column information. These commands are standard for a data integration and enrichment platform.
[INDIRECT_PROMPT_INJECTION]: The skill processes table data fetched from an external source (Clay's API). While this data could theoretically contain malicious instructions, the skill uses structured processing with jq and clearly defined CLI flags to manage the information, which minimizes the risk of the agent executing unintended instructions found within the data.
Ingestion points: Data enters via clay tables list, clay tables rows, clay tables get, and clay tables query commands as seen in SKILL.md.
Boundary markers: None explicitly defined in the prompts, but commands return structured JSON data.
Capability inventory: Uses the clay CLI for network reads and uses an external Write tool for local file creation (noted in the CSV export section).
Sanitization: The skill uses jq filters to extract specific fields and tokens, effectively selecting only the data needed for analysis.

tables-cli