The outcome

Publish an agent that answers or acts within a documented boundary and fails safely when knowledge, authorization, tools, or models are uncertain.

Coze is a visual platform and API for building, evaluating, publishing, and operating AI agents, workflows, knowledge, plugins, tools, and multi-channel applications. Composing and publishing AI agents and workflows with prompts, models, knowledge, variables, plugins, APIs, and channel integrations without building every orchestration surface from scratch. This guide narrows that broad capability into one repeatable outcome, with checkpoints that keep the source material and your judgment in the loop.

Before you begin

Set the boundary before the tool starts.

Choose one real task, identify who will use the result, and decide what evidence or test will make the result acceptable. Gather only the source material needed for that task. If the work contains confidential, personal, regulated, or client-owned information, confirm that the platform and account are approved before sharing it.

Troiana principle

AI should make the work easier to inspect. If the workflow removes the source, the owner, or the review step, redesign the workflow.

Step by step

A workflow you can repeat.

  1. 01

    Define users, purpose, prohibited uses, data classes, knowledge owners, models, tools, channels, output and citation rules, side effects, budget, retention, escalation, and rollback.

  2. 02

    Create an isolated project, ingest minimal approved knowledge with source and expiry metadata, choose reviewed models and plugins, and keep prompts and variables free of secrets.

  3. 03

    Build explicit workflow inputs, branches, schemas, timeouts, retries and stopping conditions; restrict tool destinations and permissions, reauthorize resources, and require approval before consequential actions.

  4. 04

    Create evaluations for normal, ambiguous, stale, conflicting, no-answer, multilingual, unsafe, injection, knowledge-poisoning, permission, plugin-failure, duplicate-action and token-exhaustion cases.

  5. 05

    Gate publication on results, publish privately to a limited audience, monitor traces, feedback, cost and tool calls, refresh reviewed knowledge, revoke compromised integrations, and retain a prior version for rollback.

Working standard

What good use looks like.

  • Use evaluations before publication.
  • Require evidence or abstention.
  • Approve consequential tool calls.

Coze agents can combine third-party models, plugins, APIs, knowledge, and publishing channels, so every component adds separate data terms and trust boundaries. Coze warns against exposing personal access tokens in browser code, places end-user notice and consent duties on developers, and requires lawful handling of plugin requests and protected information. Minimize shared data, review every third party, scope credentials, validate and reauthorize tool calls, make writes idempotent, require approval for consequential actions, publish privately first, and support correction, export, revocation, and deletion.

Official references

Check the current product documentation.

Features, plan limits, availability, and data controls change. These official pages are the starting points used for this collection.