Directory Researcher
Structured web data collection
Pulls a structured dataset out of a public directory, marketplace or review site. The schema is agreed up front and normalised as it goes — consistent country names, cleaned URLs, standardised sizes — so what you get is ready to work rather than ready to clean. It stays inside what's public and permitted, and reports when a source blocks it rather than evading the block.
Connects with
Connect these once. Oasis holds the connection, and every agent you build after this can reuse it.
Google SheetsRequired
FirecrawlOptional
ExaOptional
- +Custom toolConnect any other tool from thousands of available integrations.
Skills it ships with
How it runs
- Foundation
OpenAI Agents
- Model
- GPT-5.6 Terra
- Reasoning
- Medium effort
- Per run
- Up to 40 turns
- Memory
- Off
- Browsing
- Can read the web
- Category
- Sales & Marketing
- Scraping
- Data
- Research
Frequently asked questions
Can Directory Researcher turn a review site into a spreadsheet-ready dataset?
Yes. It collects records from a public directory, marketplace or review site into an agreed schema and normalises fields such as country names, URLs and sizes as it goes.
How do I define fields for Directory Researcher before a collection run?
Agree the schema up front, including the fields and target format you need. The agent uses that definition while collecting and normalising public records.
What does Directory Researcher do when a public source blocks collection?
It reports the block and stays within public, permitted access. It does not evade the source's restriction.
Will Directory Researcher drop listings with values it cannot normalise?
The workflow is designed to return a dataset ready to work with, while keeping collection inside the agreed schema. Any source limitation is reported rather than hidden; the catalog does not promise silent recovery of missing values.
Agents like this one
Put Directory Researcher to work today
Add it to your Oasis, review its setup, and hand it the first task.