The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local CLI commands using the 'revos' tool (revos cubes meta and revos cubes query) to fetch metadata and execute queries against a semantic model.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from external tool outputs and local YAML configuration files (cubes/*.yml), presenting an attack surface where untrusted data could potentially influence agent behavior.
Ingestion points: Data is ingested from the output of the revos cubes meta and revos cubes query commands, as well as the content of project files in the cubes/ directory.
Boundary markers: The skill uses explicit instructions to format the output as ASCII tables and charts, providing a clear structure for the processed data.
Capability inventory: The skill has the capability to execute the revos CLI tool, read local YAML configuration files, and write temporary JSON files to /tmp/q.json for query execution.
Sanitization: The skill relies on the agent to format and round data for display purposes; no explicit sanitization or filtering of the raw tool output is mentioned in the instructions.

query-semantic-model