# Read a sheet or table so the agent can answer from it

> Read a sheet or table so the agent can answer from it: 7 companies fully do it, 3 with a limit, 10 carry the tag. Every listing in Spreadsheets & operational tables with the reason and a source.

- Canonical: https://www.anchorterminal.com/niches/spreadsheets.read-table/
- Markdown: https://www.anchorterminal.com/niches/spreadsheets.read-table/index.md (~1,250 tokens)
- Slim: https://www.anchorterminal.com/niches/spreadsheets.read-table/index.min.md (~280 tokens, same facts, less prose, for token-sensitive contexts)
- JSON: https://www.anchorterminal.com/niches/spreadsheets.read-table/index.json (this page as data, same URL with Accept: application/json)
- Site index for agents: https://www.anchorterminal.com/llms.txt (the key docs and every listing in one file, at most 16 MB, about 4 million tokens: https://www.anchorterminal.com/llms-full.txt)
- API: https://www.anchorterminal.com/api/v1/index.json
- MCP: https://www.anchorterminal.com/mcp (the whole catalogue as tools, no key)
- DNS: dig +short TXT <slug>.grade.anchorterminal.com (the grade as a DNS TXT record, for a sandbox with no HTTP where names still resolve; also top., best.<category>. and <a>.vs.<b>. under the zone, and the same under grade.letme.dev)
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)
- Updated: 2026-11-15

Four or more companies' docs show a full fit.

| Companies, full fit | With a limit | Carry the tag | Share of the category's work | First-hand accounts |
| --- | --- | --- | --- | --- |
| 7 | 3 | 10 | 28 | 10 |

## What counts as done

Every requested row returned as typed values with column names (or field ids), paged, with the file found by name.

Measured, for our own tests: no single response exceeds about 10,000 tokens, and the file is found in at most two calls.

How agents ask for it: what's in the jobs sheet; summarise this spreadsheet; how much did we spend on subscriptions; read the leads table

## Does the job

The docs show it meets the test above, and nothing we read says otherwise.

- **Airtable (Formagrid Inc (doing business as Airtable))**: List bases returns names and Get base schema returns table and field names and ids, so a named table is found in two calls. List records pages at pageSize up to 100 with an offset and returnFieldsByFieldId; values come back typed JSON.
- **Baserow (Baserow B.V.)**: GET /api/database/tables/all-tables/ lists every table a database token can reach by name, then list rows pages with page and size, user_field_names for column names, and {link_field}__join to pull fields from linked rows.
- **Coda (Superhuman Docs) (Superhuman Platform Inc. (parent of Coda Project LLC))**: GET /docs takes a query to find a doc by name and GET /docs/{id}/tables lists tables, then rows page with limit and pageToken, with useColumnNames and valueFormat for typed values. X-Coda-Doc-Version: latest guards against a stale snapshot.
- **Microsoft Excel (Microsoft Graph workbook API) (Microsoft Corporation)**: drive/root/search(q=...) finds the workbook by name, then GET .../tables/{name}/rows pages with $top and $skip, and columns come from the table header; values are JSON cell values.
- **NocoDB (NocoDB Inc (doing business as NocoDB))**: List bases and List tables return titles, then List Table Records pages with page, pageSize or limit and offset (100 a page on Cloud), with fieldIdOnResult for stable field ids and nested linked records up to nestedLimit.
- **SeaTable (SeaTable GmbH)**: A Base-Token is bound to one base, Get Metadata returns its tables and columns by name, and List Rows takes table_name, view_name, start and limit; the SQL endpoint returns typed rows with column metadata.
- **Smartsheet API + MCP (Smartsheet Inc.)**: GET /search with query or GET /sheets finds a sheet by name, then GET /sheets/{id} pages with page and pageSize, narrows with columnIds, and returns typed cells keyed by column id with titles in the column list.

## Does the job, with a limit

It does the job, but something in the docs or a user's account will get in an agent's way. The limit is named in each row.

- **Google Sheets API (Google LLC)**: Files open by id only; finding one by name needs Drive files.list with name contains, and acc-11's agent couldn't find a sheet without it. values.get has no row paging, so large tables need split A1 ranges; acc-28 shows merged cells carry only the first value.
- **Grist (Grist Labs Inc.)**: GET /orgs/{org}/workspaces lists documents by name and GET /docs/{id}/tables lists tables, then /records returns typed values by column id. Records take limit but no offset (acc-18, open), so paging means falling back to the SQL endpoint with LIMIT and OFFSET.
- **Teable (Teable AI, Inc.)**: GET /space/{id}/search finds bases and tables by name, and list records pages with take (up to 1,000) and skip or a cursor, with fieldKeyType=id. acc-23 (open) shows link fields return only id and title, so linked values cost extra calls.

## What stops agents finishing

- the file can only be opened by id and the user gives a name
- no paging, so a large table floods context
- token reaches every file the user owns

## Around it

Before: file-storage.find-file. After: data.analyse, document-extraction.summarise.

## Where this comes from

Weight 28 of 100 in Spreadsheets & operational tables: 10 of 35 accounts; the most common job across HN (finance sheets, pull-and-push CLI, Grist via API) and the Grist and Teable read issues

As data: https://www.anchorterminal.com/api/v1/use-cases/spreadsheets.read-table.json, or the MCP tool get_use_case with id spreadsheets.read-table. Draft, read 2026-11-15 from public docs and first-hand accounts, not yet reviewed by a person.

---

Query this catalogue directly over MCP: https://www.anchorterminal.com/mcp (Streamable HTTP, no key). How: https://www.anchorterminal.com/agents/#mcp
