code-reasoning

@mettamatt/code-reasoning is a local Model Context Protocol (MCP) server (TypeScript) meant to help coding-focused assistants perform structured, step-by-step “sequential thinking” for complex programming tasks. It also provides ready-to-use prompt templates/commands for MCP clients such as Claude Desktop and related tools.

Evaluated Mar 30, 2026 (21d ago)
Repo ↗ Ai Ml mcp model-context-protocol code-assistant reasoning prompts typescript claude-desktop developer-tools
⚙ Agent Friendliness
58
/ 100
Can an agent use this?
🔒 Security
33
/ 100
Is it safe for agents?
⚡ Reliability
29
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
78
Documentation
65
Error Messages
0
Auth Simplicity
95
Rate Limits
10

🔒 Security

TLS Enforcement
50
Auth Strength
10
Scope Granularity
5
Dep. Hygiene
50
Secret Handling
60

The tool appears to be a local MCP server launched by the user (via npx) with no documented auth. Security depends largely on the MCP client/model environment and local execution. README mentions a debug mode and a safety limit of 20 thought steps, but does not describe data handling, logging redaction, or threat model details.

⚡ Reliability

Uptime/SLA
0
Version Stability
55
Breaking Changes
30
Error Recovery
30
AF Security Reliability

Best When

You want an MCP tool that improves coding reasoning flow in a desktop/client environment and you can run/install the server locally (e.g., via npx).

Avoid When

You need a networked, authenticated REST/GraphQL service, or you require well-specified operational SLOs, SLAs, and rate-limit contracts from the provider.

Use Cases

  • • Assisting with complex software design and debugging using structured step-by-step reasoning
  • • Guiding code analysis/refactoring by breaking work into multiple reasoning steps
  • • Providing templated prompts for common development tasks via MCP client UI/commands
  • • Interactive coding support workflows in Claude Desktop / MCP-capable clients

Not For

  • • Using as a general-purpose agent for non-coding domains
  • • A production hosted API service requiring centralized auth/rate limiting
  • • Any workflow requiring strong guarantees about prompt safety beyond stated “thought step” loop prevention

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
No
Webhooks
No

Authentication

Methods: None described (runs locally via npx).
OAuth: No Scopes: No

README describes local MCP server configuration and CLI flags only; no authentication mechanism is documented.

Pricing

Free tier: No
Requires CC: No

No pricing information was provided in the supplied content.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • ⚠ MCP server behavior may depend on the MCP client’s integration (Claude Desktop / VS Code).
  • ⚠ No explicit guidance on retry/error handling semantics for tool calls is provided in the supplied README content.

Alternatives

Any MCP server that provides structured reasoning templates for coding (similar “reasoning” MCP tools) General code-assistant frameworks or local LLM toolchains with custom prompting Sequential/thought-structured prompting approaches implemented directly in your own MCP/tool wrapper

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for code-reasoning.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo
API endpoint ↗ Agent guide ↗ Report inaccuracy

Scores are editorial opinions as of 2026-03-30.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
♥
Community Powered