Monday.com Accountability

Manage accountability items on the configured Monday.com board. Use when creating new accountability items, checking on existing ones, running work sessions,...

Key points to verify before installing: - Configuration mismatch: The SKILL.md names plugin config keys (mondayApiToken, boardId) but the helper script expects MONDAY_API_TOKEN and MONDAY_BOARD_ID environment variables or a token in ~/.openclaw/.env. Ask the author which is correct and update the manifest accordingly. - Credentials: The script needs your Monday.com API token and board ID. Only provide a token with the minimum necessary scope (prefer board-scoped or limited permissions) and avoid supplying a full-admin token. - Local .env access: The script will try to grep ~/.openclaw/.env for MONDAY_API_TOKEN — confirm you’re comfortable with the skill reading that file and that the file does not contain other sensitive tokens. - Binaries: The script uses curl and jq but the skill doesn't declare required binaries. Ensure jq is available in the runtime environment or request the author to declare it in the manifest. - Broad agent actions: The SKILL.md allows spawning sub-agents, performing code-related work, and messaging people. Confirm limits on who may be messaged and whether spawning other agents is allowed in your deployment. Consider running in a sandboxed/test board first. - Test safely: Try the skill with a test token and test board, monitor the API calls, and review any automated updates the agent writes. If you need higher assurance, ask the author to: (1) fix config/env mismatch, (2) declare required binaries, and (3) document exact permissions and messaging endpoints.

Type: OpenClaw Skill Name: monday-accountability Version: 0.1.0 The skill instructions in SKILL.md create a significant indirect prompt injection surface by mandating that the agent read and strictly follow instructions retrieved from a remote Monday.com board's 'Details' column. This risk is amplified by instructions to spawn sub-agents (like Claude Code) to execute work based on this potentially untrusted remote data. Additionally, the `monday-api.sh` script accesses sensitive credentials from the user's home directory (~/.openclaw/.env), which, while functional, represents a high-privilege behavior.