Companies
Tools for finding companies, reading their DealTraq record, gathering research context, and reading research saved with earlier votes.
These tools find companies and assemble what DealTraq knows about them. All of them are read-only.
search_companies
Read · Cost 2
Finds companies whose name, website, description, source, or founder names contain your search text (case-insensitive). Results are sorted by name and include each company's stage.
| Input | Type | Required | Description |
|---|---|---|---|
query | string, up to 500 characters | Yes | Text to look for. |
limit | integer, 1 to 20 | No | Maximum results. Defaults to 10. |
Notes
- This is a text match, not a pipeline query. To list everyone in a stage,
your assistant uses
list_companies_in_pipeline_stage.
Example: Find Acme in DealTraq.
get_company
Read · Cost 5
Returns a company's current record: description, website and links, stage,
lead, team assignments, founders, primary contacts, source, custom fields, and
when it was last updated. It does not read files or meetings; for research,
your assistant uses get_company_research_context.
| Input | Type | Required | Description |
|---|---|---|---|
companyId | UUID | Yes | The company to read. Your assistant gets this from a search or stage list. |
Notes
- The returned
updatedAtis needed by write tools that protect against overwriting newer changes, such aspropose_company_changes.
Example: Who is the lead on Acme and what stage is it in?
get_company_research_context
Read · Cost 7
The starting point for diligence on one company. Returns the company record plus a complete inventory of its files and meetings, a ready-made read request for every readable file and transcript, and explicit gaps for anything that cannot be read. It also lists up to 10 linked Research Scout programs and points to any research saved with earlier votes.
| Input | Type | Required | Description |
|---|---|---|---|
companyId | UUID | Yes | The company to research. |
artifactLimit | integer, 1 to 50 | No | Files and meetings per page. Defaults to 20. |
artifactCursor | integer | No | Page position. Set by nextRequest. |
Notes
- Follow
nextRequestuntil the inventory'scoverage.completeistrue, then follow each entry inevidenceReadRequests. - Unreadable items (for example, files stored with an external provider, or
meetings without a transcript) are listed with an
unavailableReasonrather than silently skipped. discoveryProgramRequestcontinues the Research Scout program list;retainedResearchRequestopens saved vote research.
Example: Review Acme's deck and latest meetings, and list the unresolved diligence questions.
list_company_research_submissions
Read · Cost 3
Lists research reports that were saved together with Inbox votes on a
company, newest first. Each item includes when it was submitted, the vote it
supported, its number of sources, and a readRequest to open the full report.
These reports were written by AI agents and are not verified by DealTraq.
| Input | Type | Required | Description |
|---|---|---|---|
companyId | UUID | Yes | The company whose saved research to list. |
limit | integer, 1 to 20 | No | Items per page. Defaults to 10. |
afterId | UUID | No | Page position. Set by nextRequest. |
Example: What research was submitted with earlier votes on Acme?
read_company_research_submission
Read · Cost 5
Reads one saved research report in full: the original Markdown report, its cited sources and excerpts, the gaps it listed, and who submitted it through which AI client. Any summaries are returned separately. Everything in the report is unverified; this tool does not fetch the cited URLs or change any company data or notes.
| Input | Type | Required | Description |
|---|---|---|---|
companyId | UUID | Yes | The company the report belongs to. |
submissionId | UUID | Yes | The report to read, from list_company_research_submissions or a vote result. |
Example: Open the research behind my upvote on Acme and check which claims are unsourced.