flowfi_skill

This skill provides guidance and actionable steps for FlowFi REST API workflows, including authentication, generate, edit, deploy, and draft management.

Python

2.6k

GitHub Stars

3

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill openclaw/skills --skill flowfi

_meta.json625 B
README.md6.4 KB
SKILL.md1.9 KB

Overview

This skill provides REST API instructions for FlowFi workflow operations: authorization, smart accounts, AI-driven generation and editing, deployment, undeployment, and draft management. It’s designed for backend integration or when answering questions about workflow and auth endpoints. Use the endpoints with a JWT bearer token for protected routes.

How this skill works

The skill documents the HTTP endpoints, required headers, request bodies, and typical responses for common FlowFi operations. It explains how to obtain and use a long-lived bearer token, list and inspect smart accounts, generate and edit workflows with natural-language prompts, get AI suggestions, and transition workflows between draft and active states. Ownership and authorization are enforced via the JWT.

When to use it

Integrating a backend or service with FlowFi workflow APIs
Building a UI that generates or edits workflows via prompts
Automating workflow deployment and lifecycle operations
Querying or listing smart accounts and workflow counts
Explaining auth/bearer token usage and token revocation

Best practices

Always send Authorization: Bearer <jwt> for protected routes and verify token expiry before calls
Use GET /smart-accounts to pick a valid smartAccountId when generating workflows
Undeploy a workflow before editing it with POST /ai/workflow/:id/prompt if it’s currently deployed
Validate responses for required fields (id, nodes, connections) after POST /ai/generate-workflow
Handle 400 errors from deploy/undeploy calls and surface actionable messages to users

Example use cases

Create a draft workflow from a user prompt: POST /ai/generate-workflow with prompt and smartAccountId
Offer inline AI edit suggestions: GET /ai/workflow/:id/suggestions to populate UI suggestions
Apply a user instruction to an existing draft: POST /ai/workflow/:id/prompt to add delays or notifications
Deploy an approved workflow to make it active: POST /workflows/:id/deploy
Revoke programmatic bearer tokens for maintenance: POST /auth/bearer-token/revoke (requires JWT)

FAQ

Exchange your app JWT for a long-lived bearer token via POST /auth/bearer-token (include Authorization: Bearer <jwt>). The response is a new JWT you can use for API clients.

Can I edit a deployed workflow with AI prompts?

No. A workflow must be draft or ended to be edited. Undeploy it first with POST /workflows/:id/undeploy, then call POST /ai/workflow/:id/prompt.

What fields are required when generating a workflow?

POST /ai/generate-workflow requires prompt (string) and smartAccountId (string). The endpoint returns a draft workflow object with id, name, nodes, and connections.