mcpx

mcpx is a command-line tool/skill that turns MCP (Model Context Protocol) servers into composable CLI commands. It can list servers/tools, call MCP tools by name, cache results, discover existing MCP server configs from other clients, and manage local config (config.toml), shims, and agent guidance blocks/skills.

Evaluated Mar 30, 2026 (23d ago)
Repo ↗ DevTools devtools mcp cli agent-workflows shell automation tooling
⚙ Agent Friendliness
64
/ 100
Can an agent use this?
🔒 Security
55
/ 100
Is it safe for agents?
⚡ Reliability
30
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
55
Documentation
80
Error Messages
--
Auth Simplicity
70
Rate Limits
10

🔒 Security

TLS Enforcement
70
Auth Strength
55
Scope Granularity
30
Dep. Hygiene
50
Secret Handling
70

Security posture is inferred from docs: mcpx itself avoids OAuth flows and does not store third-party credentials, but it supports injecting auth material via config env/header (e.g., tokens) so safe handling depends on users/agents and the underlying MCP servers. Rate limits and security/privacy guarantees are not documented in the provided content. TLS enforcement for outbound MCP endpoints is not explicitly stated, though typical CLI behavior is HTTPS-compatible.

⚡ Reliability

Uptime/SLA
0
Version Stability
45
Breaking Changes
35
Error Recovery
40
AF Security Reliability

Best When

You want to let an agent reliably invoke MCP tool schemas through a deterministic CLI interface, optionally with caching and machine-readable outputs.

Avoid When

You need strict governance around tool execution/sandboxing or require a hosted, multi-tenant API with formal SLAs.

Use Cases

  • Enable an agent to call MCP tools via straightforward CLI commands
  • Create repeatable workflows by composing MCP tool calls with shell pipelines (including JSON output)
  • Discover and reuse MCP server configs from other tools (Cursor/Claude Code/Cline/etc.)
  • Bootstrap new MCP server configs from endpoints or manifests (mcpx add)
  • Speed up repeated MCP tool calls via caching
  • Expose MCP-connected apps as servers in an agent when Codex Apps are enabled locally

Not For

  • As a full replacement for writing and running MCP servers themselves
  • Interactive OAuth-based integrations for third-party services (mcpx does not run OAuth flows)
  • Use as a hosted API/service for external users (it’s primarily a local CLI/daemon tool)

Interface

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

Authentication

Methods: MCP server authentication via config (e.g., headers/env like GITHUB_TOKEN) Optional managed guidance/skill installation for an agent (not third-party auth) Codex Apps authentication handled by Codex locally; mcpx re-exposes connected apps without running OAuth
OAuth: No Scopes: No

Authentication is delegated to the underlying MCP servers / linked services via environment variables or headers in mcpx config. mcpx itself does not run OAuth flows or store third-party credentials.

Pricing

Free tier: No
Requires CC: No

No pricing information in the provided README content.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Tool output is mostly pass-through; agents must handle provider-specific formats and parse only when --json is available for mcpx-owned surfaces.
  • Caching behavior depends on --cache/--no-cache and per-server default_cache_ttl; agents should be explicit when freshness matters.
  • Shims/skills and guidance writes modify local/global agent configuration files; agents should avoid unintended overwrites.

Alternatives

Full Evaluation Report

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

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

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

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