For Agents
Create human-in-the-loop review tasks, register reviewer agents, and pull the next assignment via round-robin distribution.
Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the EnforcedFlow API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.
Step 1: Jentic One Host machine
# On the machine that will host your Jentic One instance:
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fenforcedflow.com%2Fenforcedflow" | shStep 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fenforcedflow.com%2Fenforcedflow" | sh
jentic register # connects your agent to your Jentic One instanceJentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.
What an agent can do with EnforcedFlow API.
Create a human-in-the-loop review task tied to an automated decision
Register or update an agent within a named review group
List agents that belong to a specific review group
Fetch the next agent in a round-robin to assign work to
GET STARTED
Use for: I need to create a human-in-the-loop review task, Add a new reviewer agent to my approvals group, Find all agents in the 'compliance' group, Get the next reviewer to assign a task to
Not supported: Does not handle full workflow design, document signing, or chat-based collaboration - use for creating human-in-the-loop tasks and round-robin reviewer assignment only.
Jentic publishes the only available OpenAPI specification for EnforcedFlow API, keeping it validated and agent-ready. EnforcedFlow provides no-code AI automation with human-in-the-loop review tasks and round-robin agent management. The API lets developers create review tasks, register agents into groups, fetch the next available agent, and run simple round-robin assignment for any list of items. It is targeted at teams that want AI workflows where a human reviews or approves automated decisions.
Run round-robin assignment for arbitrary items, not just agents
Patterns agents use EnforcedFlow API for, with concrete tasks.
★ AI Output Review Queue
When an AI agent produces a sensitive output (e.g. a contract clause or a refund decision), it calls POST /tasks to file a human-in-the-loop review task. EnforcedFlow then routes that task to the next available reviewer. This gives teams a quick way to add an oversight layer to LLM workflows without building a custom queue.
On any LLM output flagged 'high-risk', create a review task in EnforcedFlow with the relevant context
Reviewer Pool Management
Use POST /agents to register a reviewer into a named group and GET /agents to list everyone currently on rotation. Combined with GET /agents/next, this gives ops teams a working assignment queue with one API surface and no infrastructure of their own.
Register reviewer 'alice@example.com' in group 'compliance' and confirm she appears in the group listing
Generic Round-Robin Distribution
GET /round-robin/next returns the next item from any list using a simple round-robin counter, useful for distributing leads, tickets, or escalations beyond just reviewer assignment. The same primitive backs the agent assignment flow.
Call the round-robin endpoint to assign the next inbound lead to one of three sales reps
Agent-Driven Approval Pipeline
An AI agent that drafts customer-facing replies searches Jentic for 'create a review task', loads EnforcedFlow's POST /tasks schema, and creates a task each time the draft passes a confidence threshold for human review. Reviewers are pulled via GET /agents/next so assignment stays balanced.
On every drafted email above 200 words, create a task in EnforcedFlow and assign it to the next agent in 'email-review'
5 endpoints — jentic publishes the only available openapi specification for enforcedflow api, keeping it validated and agent-ready.
METHOD
PATH
DESCRIPTION
/tasks
Create a human-in-the-loop task
/agents
Add or update an agent
/agents
List agents in a group
/agents/next
Get the next agent in round-robin
/round-robin/next
Get the next item in a generic round-robin
/tasks
Create a human-in-the-loop task
/agents
Add or update an agent
/agents
List agents in a group
/agents/next
Get the next agent in round-robin
/round-robin/next
Get the next item in a generic round-robin
What agents get from Jentic-routed access to this vendor.
Setup
Wiring the EnforcedFlow API by hand means setting up its X-API-Key header, formatting task, agent, and round-robin calls, and handling your own errors against https://api.enforcedflow.com/v1. Through Jentic you install once, import the EnforcedFlow API from the API Directory, store the key once, and your agent calls it.
Permission scoping
The EnforcedFlow API creates tasks and agents through body-based POST calls with no resource id in the URL path, so scope it by operation: limit the agent to the operations it needs, such as reading the next round-robin reviewer, and leave task creation out of the allowed set unless you add it.
Credential isolation
Your EnforcedFlow API key is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
Intent-based discovery
Agents search Jentic by intent such as 'create a human-in-the-loop task' or 'get the next reviewer in the rotation', and Jentic returns the matching EnforcedFlow operation with its input schema so the agent calls the right endpoint without browsing the reference docs.
Alternatives and complements available in the Jentic catalogue.
Specific to using EnforcedFlow API through Jentic.
Why is there no official OpenAPI spec for EnforcedFlow API?
EnforcedFlow does not publish an OpenAPI specification publicly. Jentic generates and maintains this spec so that AI agents and developers can call EnforcedFlow API via structured tooling. It is validated against the live API and kept up to date. Get started with Jentic One, the self-hosted execution layer.
What authentication does the EnforcedFlow API use?
The API uses an API key via the `apiKey` security scheme. Through Jentic, the key is stored encrypted in the vault and supplied at execution time so the agent never sees the raw value.
Can I create human-in-the-loop tasks with EnforcedFlow?
Yes. POST /tasks creates a review task that EnforcedFlow then routes to the next available reviewer. Use POST /agents first to register the reviewers in the relevant group.
How does round-robin assignment work in EnforcedFlow?
GET /agents/next returns the next reviewer for a group, and GET /round-robin/next returns the next item in a generic round-robin. Both endpoints maintain the rotation state server-side, so callers don't have to track who's up next.
What are the rate limits for the EnforcedFlow API?
The OpenAPI spec does not declare rate limits. Treat as workspace-scoped throttling and back off on HTTP 429; confirm the exact limits with EnforcedFlow support.
How do I create a review task through Jentic?
Run `pip install jentic`, search 'create a human in the loop review task', and Jentic returns POST /tasks with its input schema. The agent supplies task context and executes against base URL https://api.enforcedflow.com/v1.
Can I limit what my agent is allowed to do with the EnforcedFlow API?
Yes. Because you run Jentic One yourself, your own rules decide which EnforcedFlow operations the agent may call, and the API scopes cleanly by operation since tasks and agents are created through body-based POST calls with no resource id in the URL path. You can, for example, allow only read operations like GET /agents/next or GET /round-robin/next so the agent can fetch the next reviewer, while leaving POST /tasks and POST /agents out of the allowed set unless you explicitly grant them. The stored API key is injected at execution time by your instance, so the agent acts only within the operations you permit.