The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs users to add https://rube.app/mcp as an MCP server. This connects the agent to an external service that defines the tool schemas and controls the logic for Coda interactions. This introduces a dependency on a third-party service that is not part of the standard trusted vendor list.
[DATA_EXFILTRATION]: The skill provides tools with high-risk capabilities that could be exploited for data exfiltration. Specifically, CODA_ADD_PERMISSION allows granting document access to external emails, CODA_PUBLISH_DOC makes private documents publicly accessible via a URL, and CODA_BEGIN_CONTENT_EXPORT generates downloadable exports of document content.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the data it processes.
Ingestion points: Document data is ingested via CODA_LIST_TABLE_ROWS, CODA_SEARCH_ROW, CODA_SEARCH_DOCS, CODA_LIST_PAGES, and CODA_GET_A_PAGE (SKILL.md).
Boundary markers: None identified. There are no instructions to the agent to ignore or delimit instructions found within the document data.
Capability inventory: The skill has access to sensitive operations including permission management (CODA_ADD_PERMISSION), document publishing (CODA_PUBLISH_DOC), and content export (CODA_BEGIN_CONTENT_EXPORT).
Sanitization: None identified. Data retrieved from Coda is passed directly into the agent's context without filtering or escaping.
[COMMAND_EXECUTION]: The skill facilitates the execution of API commands through the Rube MCP server. The instructions require the agent to call RUBE_SEARCH_TOOLS and RUBE_MANAGE_CONNECTIONS to initialize the environment, delegating command authority to the external MCP endpoint.

coda-automation