Web research: search, read, and compare public pages
Hub for Web research guides on how the Agent searches and reads public pages, requests typed results, and troubleshoots blocked or stale sources.
This page covers Web research capabilities, terminology distinctions, and related guides for Agent public-page research.
- Web research capabilities and boundaries.
- Web search versus Web research terminology.
- Related guides for Web research workflows.
Summaries were generated by AI.
During a Task, the Agent can search and read public pages, compare sources, and use the evidence in its answer. Use Web research when a Task needs information that can change after model training, must compare public sources, or must find and read a public document.
Your application submits a normal Task request. The SDK has no separate Web research method. The Agent decides when the Task needs public evidence. To keep source URLs in the terminal result, request them in its response model.
What Web research includes
- Capability
- Research a public question
- What happens
- A bounded brief defines the question and evidence to look for
- Boundary
- The brief can become a public query; include no secrets
- Capability
- Find public pages
- What happens
- Focused queries return candidate pages
- Boundary
- A search result helps discovery; it is not source evidence
- Capability
- Read selected pages
- What happens
- Selected pages provide extracted content, a title, and the final URL
- Boundary
- Page access is not an interactive browser or a supported sign-in flow
- Capability
- Compare source material
- What happens
- Research produces a compact source-linked synthesis for the Task
- Boundary
- A cited result still needs source review
Use precise terms
This documentation calls public-page discovery Web search. Web research is the larger workflow that investigates a question, runs searches, reads selected pages, compares sources, and synthesizes the evidence. Sources and factuality defines grounding and fact-checking.
Web search is not general Internet access. It does not provide supported access to private-network destinations or non-HTTP protocols. It also does not expose clicks, forms, credential entry, screenshots, or an interactive browser session.
Data protection defines the external paths and retention.
Treat fetched text as external input. The threat model explains how it can affect later agent work and where your integration must enforce approval.
Choose a guide
- How Web research works—follow a source review from research brief to task result.
- Request Web research—ask for dated sources, gaps, and a typed result.
- Sources and factuality—decide what a source link proves and what you must verify.
- Page access—check which page behavior and session state the agent can use.
- Troubleshoot Web research—recover from missing, stale, blocked, or unreadable sources.
Scope and limits states permitted use. The Agent harness explains how the Agent uses Tools to complete a Task.