Sequential Thinking Tools MCP server
Commit Nov 15, 2026Provides a scratchpad for recording sequential reasoning steps, branching plans, and validating model-authored tool recommendations.
Also known as spences10/mcp-sequentialthinking-tools
- Stars
- 588
- Tools
- 3
- Downloads, 30 days
- —
- Latest version
- —
- Last commit
- Nov 15, 2026
- License
- MIT
What is Sequential Thinking Tools MCP?
Sequential Thinking Tools provides a scratchpad for recording sequential reasoning steps, branching plans, and validating model-authored tool recommendations. It exposes 3 tools, such as sequentialthinking_tools, get_thinking_history, clear_thinking_history.
Summary generated from the project's README
About Sequential Thinking Tools MCP server
Recording complex reasoning steps and branching plans from inside an AI client is what this server enables. It acts as a dedicated scratchpad for developers who need to break down messy problems into inspectable, sequential steps. By using the sequentialthinking_tools command, an agent can log individual thoughts, manage session history, and create branches for different plan iterations.
This tool is particularly useful for validating model-authored tool recommendations. If you provide a list of available_tools, the server checks that any recommended_tools exist before storing the step, returning an error if names do not match. It includes security features that scan and redact prompt-injection-like text to protect the integrity of the reasoning history.
Users can manage their work through get_thinking_history to retrieve stored thoughts or clear_thinking_history to wipe specific sessions. The server runs locally via npx and can be configured with a MAX_HISTORY_SIZE environment variable. It is designed for complex tasks where explicit reasoning is required rather than trivial requests.
What you can do with it
- 1Deconstruct complex problemsBreak a messy problem into smaller, manageable steps for better clarity.
- 2Validate tool recommendationsEnsure that the tools an AI model suggests actually exist in the provided tool list.
- 3Manage branching logicCreate and track different versions of a plan using branching and revision metadata.
- 4Audit reasoning historyInspect the full history of thoughts during a long agent run to understand the decision process.
Written from the project's README. Every use case is checked against a quote from it.
How to install Sequential Thinking Tools MCP
Formats verified 10 Nov 2026Add to CursorAdd to GooseInstall Sequential Thinking Tools MCP in Claude Code
$ claude mcp add mcp-sequentialthinking-tools -- npx -y mcp-sequentialthinking-toolsInstall Sequential Thinking Tools MCP in Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"mcp-sequentialthinking-tools": {
"command": "npx",
"args": [
"-y",
"mcp-sequentialthinking-tools"
]
}
}
}Install Sequential Thinking Tools MCP in VS Code
.vscode/mcp.json
{
"servers": {
"mcp-sequentialthinking-tools": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"mcp-sequentialthinking-tools"
]
}
}
}Install Sequential Thinking Tools MCP in VS Code CLI
$ code --add-mcp '{"name":"mcp-sequentialthinking-tools","type":"stdio","command":"npx","args":["-y","mcp-sequentialthinking-tools"]}'Install Sequential Thinking Tools MCP in Claude Desktop
claude_desktop_config.json
{
"mcpServers": {
"mcp-sequentialthinking-tools": {
"command": "npx",
"args": [
"-y",
"mcp-sequentialthinking-tools"
]
}
}
}Install Sequential Thinking Tools MCP in Codex
$ codex mcp add mcp-sequentialthinking-tools -- npx -y mcp-sequentialthinking-toolsInstall Sequential Thinking Tools MCP in Codex config.toml
~/.codex/config.toml
[mcp_servers.mcp-sequentialthinking-tools] command = "npx" args = ["-y", "mcp-sequentialthinking-tools"]
Install Sequential Thinking Tools MCP in Gemini CLI
$ gemini mcp add -s user mcp-sequentialthinking-tools npx -- -y mcp-sequentialthinking-toolsInstall Sequential Thinking Tools MCP in Gemini settings.json
~/.gemini/settings.json
{
"mcpServers": {
"mcp-sequentialthinking-tools": {
"command": "npx",
"args": [
"-y",
"mcp-sequentialthinking-tools"
]
}
}
}Started by HostedMCP on 16 Nov 2026
We ran this command for mcp-sequentialthinking-tools 0.1.1 in an isolated container with no network and no keys. It answered and listed 3 tools.
Sequential Thinking Tools MCP tools 3
Read from the server with tools/listsequentialthinking_toolsRecord one step of sequential reasoning. Optionally include available_tools and recommended_tools; recommendations are validated, not generated, by this server.
get_thinking_historyReturn recorded thoughts for a session. Use this to inspect or resume prior reasoning.
clear_thinking_historyClear recorded sequential-thinking history for one session or all sessions.
| sequentialthinking_tools | Record one step of sequential reasoning. Optionally include available_tools and recommended_tools; recommendations are validated, not generated, by this server. | |
| get_thinking_history | Return recorded thoughts for a session. Use this to inspect or resume prior reasoning. | |
| clear_thinking_history | Clear recorded sequential-thinking history for one session or all sessions. |
Configuration
MAX_HISTORY_SIZE
| Name | Description | |
|---|---|---|
| MAX_HISTORY_SIZE | — |
Automated checks
Ran Nov 16, 2026Full report →Not verified yet. A server needs a score of 80, no serious finding, and enough data for us to check. How checks work · What to fix for more points
- Source repository linked and reachablespences10/mcp-sequentialthinking-tools is public and reachablegithub.com/spences10/mcp-sequentialthinking-tools ↗Pass
- Package names this repository as its sourcemcp-sequentialthinking-tools names spences10/mcp-sequentialthinking-tools as sourcenpmjs.com/package/mcp-sequentialthinking-tools ↗Pass
- Build provenance (Sigstore attestation)No build attestation published for this releaseNot checked
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-11-15github.com/spences10/mcp-sequentialthinking-tools ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNo popular package one edit away from mcp-sequentialthinking-toolsPass
- No critical advisory in the package or its dependenciesNot checked
- No hidden instructions to the model3 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNo tool named like delete, drop or execPass
- Remote endpoint answers the MCP handshakeNot checked
README
View on GitHubmcp-sequentialthinking-tools
A lightweight MCP server for recording sequential reasoning steps. It is a scratchpad with history, branching, revision metadata, and optional validation for model-authored tool plans.
It does not discover your other MCP tools and it does not
choose tools for the model. If you pass available_tools and
recommended_tools, the server validates that the recommended names
exist and stores the step.
Why use it?
Use this when a task benefits from explicit, inspectable reasoning:
- breaking a messy problem into steps;
- revising or branching a plan;
- keeping a small reasoning history by session;
- validating tool-plan names against a supplied tool list;
- clearing or inspecting reasoning history during a long agent run.
Do not use it for trivial requests. It adds overhead.
Tools
sequentialthinking_tools
Records one thought.
Required parameters:
thought— current reasoning stepthought_number— current step numbertotal_thoughts— current estimate; automatically raised if lower thanthought_numbernext_thought_needed— whether another thought is needed
Optional parameters:
session_id— history bucket; defaults todefaultis_revision,revises_thoughtbranch_from_thought,branch_idneeds_more_thoughtsavailable_tools— array of tool names or{ name, description }objectsrecommended_tools— model-authored recommendations to validate/storeremaining_steps— short list of upcoming steps
Example:
{
"session_id": "svelte-debug",
"thought": "First inspect the route files, then run the failing check.",
"thought_number": 1,
"total_thoughts": 3,
"next_thought_needed": true,
"available_tools": ["read", "bash"],
"recommended_tools": [
{
"tool_name": "read",
"confidence": 0.9,
"rationale": "Need to inspect the relevant files before editing.",
"priority": 1
}
]
}
If recommended_tools contains a name not present in
available_tools, the call returns isError: true and does not store
the thought.
Security posture
The server treats thought text, tool descriptions, rationales, and
remaining-step text as untrusted input. Prompt-injection-like text is
scanned and redacted before it is stored or returned in history. Calls
with redactions include security_warnings showing which fields
matched.
This is defensive filtering, not a guarantee that arbitrary adversarial text is safe. Do not put secrets in thoughts or tool descriptions.
get_thinking_history
Returns stored thoughts for a session.
Parameters:
session_id— defaults todefaultbranch_id— optional branch filterlimit— max records to return; default50, max500
clear_thinking_history
Clears one session or every session.
Parameters:
session_id— defaults todefaultall_sessions— clear all history buckets
Prompt
sequential-thinking-guidance
A short prompt that tells the model how to use this server honestly: as a scratchpad and validator, not as an external reasoning engine.
Configuration
Claude Desktop / compatible MCP clients
{
"mcpServers": {
"mcp-sequentialthinking-tools": {
"command": "npx",
"args": ["-y", "mcp-sequentialthinking-tools"],
"env": {
"MAX_HISTORY_SIZE": "1000"
}
}
}
}
MAX_HISTORY_SIZE is per session and defaults to 1000.
The server uses tmcp and includes a small stdio transport that
accepts both standard Content-Length framed MCP messages and
newline-delimited JSON used by older tmcp tooling.
Development
pnpm install
pnpm test
pnpm build
pnpm check
This project uses vite-plus for build, test, format, and lint
orchestration.
Publishing
pnpm changeset
pnpm changeset version
pnpm release
License
MIT License — see LICENSE.
Acknowledgments
- Built on the Model Context Protocol
- Adapted from the MCP Sequential Thinking Server
Sequential Thinking Tools MCP questions
How do I add Sequential Thinking Tools MCP to Claude Code?
Run claude mcp add mcp-sequentialthinking-tools -- npx -y mcp-sequentialthinking-tools in your terminal. Add --scope user to make it available in every project.
How do I add Sequential Thinking Tools MCP to Cursor?
Open ~/.cursor/mcp.json and add the server under mcpServers, using the Cursor config shown in the install section above.
What tools does Sequential Thinking Tools MCP provide?
It exposes 3 tools, including sequentialthinking_tools, get_thinking_history, clear_thinking_history.
Does this server choose tools for the model?
No, the server does not discover or choose tools; it only validates recommended tool names against a provided list.
How does the server handle security and prompt injection?
The server scans and redacts prompt-injection-like text before storage, providing security warnings if redactions occur.
Can I clear the reasoning history?
Yes, you can clear a specific session or all sessions using the clear_thinking_history tool.
What happens if a recommended tool is not in the available tools list?
The call will return an error and the thought will not be stored.