# Revise an earlier step or explore an alternative branch

> Revise an earlier step or explore an alternative branch: 0 companies fully do it, 1 with a limit, 2 carry the tag. Every listing in Reasoning scaffolds with the reason and a source.

- Canonical: https://www.anchorterminal.com/niches/reasoning.revise-and-branch/
- Markdown: https://www.anchorterminal.com/niches/reasoning.revise-and-branch/index.md (~650 tokens)
- Slim: https://www.anchorterminal.com/niches/reasoning.revise-and-branch/index.min.md (~230 tokens, same facts, less prose, for token-sensitive contexts)
- JSON: https://www.anchorterminal.com/niches/reasoning.revise-and-branch/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

No company's docs show a full fit yet. The listings below do the job with a limit, or not at all.

| Companies, full fit | With a limit | Carry the tag | Share of the category's work | First-hand accounts |
| --- | --- | --- | --- | --- |
| 0 | 1 | 2 | 15 | 4 |

## What counts as done

A revision call names the thought it replaces and a branch call names its parent, the server returns the branch list, and history from an unrelated earlier problem is not mixed in.

How agents ask for it: that step was wrong, go back; try the other approach too; compare two ways of doing this

## 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.

- **Sequential Thinking (MCP reference server) (Model Context Protocol, a Series of LF Projects, LLC)**: isRevision with revisesThought and branchFromThought with branchId, and every result lists branches. History lasts the whole connection with no reset tool, so earlier problems mix in, and acc-07 never saw a model branch.

- not this job: Clear Thought 1.5 (Waldzell AI): isRevision, revisesThought, branchFromThought and branchId are accepted in parameters and stored, but the result returns only the last three thought numbers, never a branch list, and the store keeps no branches.

## What stops agents finishing

- model never calls the branch feature
- history from earlier problems in the same connection

## Around it

Before: reasoning.plan-before-acting. 

## Where this comes from

Weight 15 of 100 in Reasoning scaffolds: 4 of 28 accounts, mostly saying branching goes unused or state is not what it seems

As data: https://www.anchorterminal.com/api/v1/use-cases/reasoning.revise-and-branch.json, or the MCP tool get_use_case with id reasoning.revise-and-branch. 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
