The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from Monday.com boards and items which are untrusted external sources.
Ingestion points: MONDAY_LIST_BOARD_ITEMS, MONDAY_ITEMS_PAGE, and MONDAY_LIST_SUBITEMS_BY_PARENT in SKILL.md.
Boundary markers: Absent; the skill does not instruct the agent to use delimiters or ignore instructions embedded in the board data.
Capability inventory: The agent can modify boards, create items, and update column values across the Monday.com workspace.
Sanitization: Absent; no filtering or validation logic is defined for the content retrieved from the API.
[EXTERNAL_DOWNLOADS]: The setup instructions direct the user to connect to an external third-party MCP server at https://rube.app/mcp which is not a pre-approved or well-known tech service organization.
[METADATA_POISONING]: The YAML frontmatter includes a non-standard risk: critical field which could be misleading or used to bypass automated categorization by imitating internal security labels.
[DYNAMIC_EXECUTION]: The tool MONDAY_CREATE_OBJECT allows for the execution of raw GraphQL mutations, providing a mechanism for dynamic query assembly that could be abused if the agent interprets malicious input as part of a mutation string.

monday-automation