openai · OpenAI Platform Docs
Agents Guardrails and human review
Teaches how to implement automated guardrails and human-in-the-loop approval workflows to control agent execution, including input, output, and tool-based validation.
Derived skill
Files assembled from official documentation
Viewing SKILL.md
Agents Guardrails and human review
Teaches how to implement automated guardrails and human-in-the-loop approval workflows to control agent execution, including input, output, and tool-based validation.
When To Use
Use when you need to implement automated validation for agent inputs and outputs or when you must pause agent execution to require manual approval for sensitive tool calls.
Reference Files
| File | Contains | Use For |
|---|---|---|
SKILL.md | Entry point: scope, routing table, and workflow. | Start here. |
docs/agents-guardrails-approvals-and-human-review-workflow-guide.md | A guide explaining how to implement automatic guardrails and human review processes for agentic workflows. | Questions about a guide explaining how to implement automatic guardrails and human review processes for agentic workflows. |
examples/agents-guardrails-approvals-and-human-review-openai-agents-guardrails.typescript | A TypeScript code example demonstrating how to implement input guardrails and agent configurations for human review workflows using the OpenAI Agents API. | Exact payloads, commands, or snippets shown in A TypeScript code example demonstrating how to implement input guardrails and agent configurations for human review w... |
examples/agents-guardrails-approvals-and-human-review-openai-agents-guardrails.python | A Python implementation demonstrating how to use input guardrails and Pydantic models to validate agent outputs for specific tasks. | Exact payloads, commands, or snippets shown in A Python implementation demonstrating how to use input guardrails and Pydantic models to validate agent outputs for s... |
examples/agents-guardrails-approvals-and-human-review-openai-agents-typescript-gu.typescript | A TypeScript example demonstrating how to implement tool-based human approval requirements within an OpenAI Agent. | Exact payloads, commands, or snippets shown in A TypeScript example demonstrating how to implement tool-based human approval requirements within an OpenAI Agent. |
examples/agents-guardrails-approvals-and-human-review-openai-agents-python-guardr.python | A Python example demonstrating how to use the @functiontool decorator with needsapproval=True to implement human-in-the-loop guardrails for an agent. | Exact payloads, commands, or snippets shown in A Python example demonstrating how to use the @functiontool decorator with needsapproval=True to implement human-in-t... |
What This Skill Covers
- Use guardrails for automatic checks and human review for approval decisions. Together, they define when a run should continue, pause, or stop.
- Main sections:
Choose the right control,Add a blocking guardrail,Pause for human review,Approval lifecycle,Workflow boundaries matter.
Workflow
- Open the most relevant file under
docs/for the exact documented workflow and wording. - Open
schemas/files for exact structured contracts. - Open
examples/files for concrete requests, commands, snippets, and manifests. - Do not add behavior or configuration that is not present in the attached source files.
Canonical source: https://developers.openai.com/api/docs/guides/agents/guardrails-approvals.md