Zapier

Complete Zapier automation with Zaps, Tables, Interfaces, webhooks, REST Hooks API, and 6000+ app integrations.

This skill is a documentation/reference wrapper for Zapier and is internally coherent. Before installing: 1) Only provide ZAPIER_API_KEY / ZAPIER_TABLES_TOKEN if you trust Zapier and this skill — prefer tokens scoped to minimal permissions and create a separate integration token if possible. 2) Confirm whether ZAPIER_TABLES_TOKEN is actually required for your use (docs call it optional but metadata marks it required). 3) Check ~/zapier/memory.md after first run and do not store long-lived secrets there; keep API keys in secure environment variables or a secret manager. 4) Review any Zaps/webhooks the agent creates and their target URLs — this skill will send mapped data to Zapier and any connected apps you configure. 5) Be aware the docs include an example chatbot 'System Prompt' — ensure your agent runtime does not automatically adopt external system prompts. 6) Test with a limited/test Zapier account and non-production data first, and rotate/revoke tokens if you stop using the skill.

Type: OpenClaw Skill Name: zapier Version: 1.0.0 The skill bundle is benign. It provides comprehensive documentation and examples for interacting with the Zapier platform. The `SKILL.md` and `setup.md` files instruct the AI agent to gather user preferences and store them locally in `~/zapier/memory.md`, which is a legitimate function for a skill to maintain state. All network interactions are directed towards official Zapier domains (e.g., `api.zapier.com`, `hooks.zapier.com`, `tables.zapier.com`). Code snippets (bash, JavaScript, Python) are presented as examples of Zapier's functionality for the user or for Zapier's internal 'Code' steps, not as direct commands for the OpenClaw agent to execute arbitrarily. There is no evidence of prompt injection attempts, data exfiltration, or other malicious intent.