The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill design involves ingesting untrusted data from the user and interpolating it into prompts for other tools, creating a potential vector for indirect prompt injection.
Ingestion points: User context regarding their role, team, company, and current projects is gathered in Step 2 of SKILL.md and through the post-connect starter prompt.
Boundary markers: The instructions do not specify the use of boundary markers or delimiters to isolate user-supplied work context when it is interpolated into the creationPrompt for the create_board tool.
Capability inventory: The skill has broad write capabilities within the monday.com environment, including create_board, update_items, create_column, create_automation, and all_monday_api (for update_group).
Sanitization: While the skill explicitly instructs the agent to avoid including personal or sensitive information in tool inputs, it lacks formal sanitization or validation logic to filter out potentially malicious instructions embedded in the user's work descriptions.

onboard-to-monday