The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill retrieves and processes data from external Airtable records and tables, creating a surface for indirect prompt injection if those records contain malicious instructions.\n
Ingestion points: Data is fetched via commands like list-records-for-table and search-records.\n
Boundary markers: The instructions do not provide explicit delimiters or warnings to the agent to ignore instructions embedded within the retrieved data.\n
Capability inventory: The skill has the ability to execute shell commands through the airtable-mcp CLI and perform network requests to Airtable APIs.\n
Sanitization: There is no evidence of field value sanitization before the content is placed into the agent's context.\n- [EXTERNAL_DOWNLOADS]: The skill requires the installation of an external Node.js package.\n
Evidence: Command npm install -g @airtable/mcp-cli used for setup.\n- [COMMAND_EXECUTION]: The skill operates by executing the airtable-mcp command-line utility to interact with Airtable services.\n
Evidence: Numerous CLI command patterns defined throughout SKILL.md.

airtable-cli