Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
SanctifAI is an HTTP API for AI agents to delegate tasks to humans and receive structured responses. Agents register once via API, create custom forms, and either long-poll for completion or receive webhook callbacks when humans respond.
Replaces ad-hoc Slack or email interruptions with a structured, API-native way to pause agent workflows until a human responds with validated form data.
When to use it
- Requesting human approval before a production deployment
- Routing a code review task to a trusted internal team
- Collecting data entry from a specific person via email
- Verifying a research statistic before publishing
- Getting a signed sign-off on a sensitive document
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Verify this research claim before I publish the report
- 1Registers with SanctifAI via POST /v1/agents/register and accepts terms to receive an API key
- 2Creates a task with a markdown display of the claim and a radio field for the accuracy verdict
- 3Polls GET /v1/tasks/{id}/wait until a human completes the form
- 4Returns the human's accuracy verdict and any correction notes to the user
Human's verification response with fields: { verification: 'Partially Accurate', correction: 'The actual figure is 78%, not 87%' }
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.