The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of the cargo-ai CLI to perform data operations, including creating models, updating columns, and executing SQL queries against the workspace storage.
[EXTERNAL_DOWNLOADS]: The skill instructions specify the installation of the @cargo-ai/cli Node.js package from the public npm registry.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the workspace when listing records or executing SQL queries, which could contain adversarial instructions.
Ingestion points: Data returned from cargo-ai storage record list and cargo-ai storage query execute (referenced in SKILL.md).
Boundary markers: Not present; the instructions do not include specific delimiters or guidelines for the agent to separate untrusted data from instructions.
Capability inventory: Ability to modify workspace resources through CLI commands like cargo-ai storage model create or cargo-ai storage column update (referenced in SKILL.md).
Sanitization: No evidence of data sanitization or filtering is provided in the skill documentation.
[CREDENTIALS_UNSAFE]: The skill provides a walkthrough for assembling a webhook URL that includes a workspace API token in the query string (?token=<api-token>). While the documentation warns against persisting this URL, it encourages the exposure of credentials in the output and via the URL parameter (referenced in references/examples/ingest-webhook.md).

cargo-storage