anthropic for Codex

8,249 mods tagged anthropic that install into Codex, measured the same four ways as everything else here: what each costs in tokens on every session, what a security scan found and why, whether it is the original, and what changed.

Also for Codex: Git and workflow · Web3 and blockchain · Python · React · TypeScript · openai · android · Go · docker · bun · PostgreSQL · Localization

agentfit

721

MukundaKatta/agentfit-mcp

MCP server Claude CodeCodexCursor +2

Token-aware message truncation: fit a chat history into your model's context budget. Runs locally from the @mukundakatta/agentfit-mcp npm package.

not rated 1 1mo ago A tokens not measured original MIT

agentvet

722

MukundaKatta/agentvet-mcp

MCP server Claude CodeCodexCursor +2

Validate tool-call args before execution. Returns LLM-friendly retry hints. Runs locally from the @mukundakatta/agentvet-mcp npm package.

not rated 1 1mo ago A tokens not measured original MIT

agentcast

723

MukundaKatta/agentcast-mcp

MCP server Claude CodeCodexCursor +2

Structured-output enforcer: extract and validate JSON from messy LLM text. Runs locally from the @mukundakatta/agentcast-mcp npm package.

not rated 1 1mo ago A tokens not measured original MIT

agentguard

724

MukundaKatta/agentguard-mcp

MCP server Claude CodeCodexCursor +2

Network-egress firewall for agent tools: check URLs against a declarative policy. Runs locally from the @mukundakatta/agentguard-mcp npm package.

not rated 1 1mo ago A tokens not measured original MIT

acr-agent-registry

725

Tethral-Inc/AgentRegistry

Skill Claude CodeCodex

Behavioral registry for AI agents. Log interactions, build an interaction profile, query it through lenses (friction, coverage, stable corridors, failure registry, trend), and receive anomaly signal notifications if ACR observes elevated signals on a component in your composition.

not rated 1 1mo ago A 56 tokens original MIT

eyeot-erp

726

Termi24/ERP-by-Eyeot-Software

MCP server Claude CodeCodexCursor +2

ERP français avec serveur MCP natif : plus de 1 800 actions métier auto-générées via OpenAPI. Remote server at erp.eyeot.fr.

not rated 1 3mo ago A tokens not measured original MIT

doxa-mcp

727

The-Doxa-Way/doxa-mcp-schema

MCP server Claude CodeCodexCursor +2

Christian AI for any question, in any season. Bible (BSB), grace + truth. faith.tools 5/5. Remote server at doxa.app.

not rated 1 11d ago A tokens not measured original MIT

bench

728

VirajMishra1/bench

MCP server Claude CodeCodexCursor +2

Discover public AI agents, reusable recipes, and trusted benchmark evidence by task. Remote server at bench.virajmishratakehome.workers.dev.

not rated 1 1mo ago A tokens not measured original MIT

mcp-memory-graph

729

YonasValentin/mcp-memory-graph

Skill Claude CodeCodex

Drive the mcp-memory-graph MCP server like an expert: store, recall, and reconcile durable knowledge across sessions over 51 local tools ($0/token, one SQLite file). Use when ANY memory tool is in play (mcpmemory), when the user says "remember this" / "recall that" / "store a decision", when a memory seems missing…

not rated 1 1mo ago A 130 tokens

opdstar-nhi

730

tatsuju/opdstar-nhi-mcp

MCP server Claude CodeCodexCursor +2

The MCP server for Taiwan NHI billing intelligence — rejection codes, audit clauses, drug payment rules, fee schedule, special materials, and appeal case precedents. Helps AI agents answer Taiwan health-insurance billing and audit questions with authorita. Runs locally from the @opdstar/nhi-mcp npm package.

not rated 1 1mo ago A tokens not measured original MIT

mcp-loom

731

buttonmonkey/mcp-loom

MCP server Claude CodeCodexCursor +2

Stdio MCP proxy: intercepts oversized tool results into a queryable embedded DuckDB. Runs locally from the mcp-loom npm package.

not rated 1 2mo ago A tokens not measured original MIT

claude-remind-mcp

732

emretheus/claude-remind-mcp

MCP server Claude CodeCodexCursor +2

Search past Claude Code conversations. Local BM25 over JSONL with one-click claude --resume. Runs locally from the claude-remind-mcp npm package. Needs 2 environment variables to run.

not rated 1 4mo ago A tokens not measured original MIT

tiktapdown-mcp

733

farukkolip/tiktapdown-mcp

MCP server Claude CodeCodexCursor +2

TikTok toolkit: download videos, niche hashtags, best posting times, viral hook formulas. Runs locally from the tiktapdown-mcp npm package.

not rated 1 2mo ago A tokens not measured

context-keeper

734

jarmstrong158/context-keeper

MCP server Claude CodeCodexCursor +2

MCP server that maintains project context (decisions, pipelines, constraints) across Claude conversations. Runs locally from the context-keeper-mcp Python package. Needs 1 environment variable to run.

not rated 1 27d ago A tokens not measured original MIT

anchor-x402

735

hypeprinter007-stack/anchor-x402-mcp

MCP server Claude CodeCodexCursor +2

MCP server for anchor-x402 — 14 x402-paid agent utilities on Base + Solana. 9 commodity primitives (anchor, screen, attest, decode/tx, resolve/name, price/token, decode/calldata, parse/datetime, intel/wallet) plus 5 universal LLM endpoints (roast, oracle. Runs locally from the anchor-x402-mcp npm package. Needs 1…

not rated 1 9d ago A tokens not measured original MIT

reddit-mcp

736

jorgen-k/reddit-mcp

MCP server Claude CodeCodexCursor +2

Local MCP server that fetches and refines Reddit (and other JSON) content for Claude, via Reddit's public RSS feeds (no auth). Runs locally from the reddit-rss-mcp Python package.

not rated 1 3d ago A tokens not measured original MIT

agent-ready

737

mlava/agent-ready-mcp

Skill Claude CodeCodex

This skill should be used when the user wants to check how readable a website is to AI agents (ChatGPT, Claude, Perplexity, Google Gemini, MCP clients). Activates for "scan this site for AI readability", "check my llms.txt", "audit our agent-readability", "is my MCP server card valid", "what's our Vercel Agent…

not rated 1 6d ago A 122 tokens original MIT

studydiff

738

nickjlamb/studydiff

MCP server Claude CodeCodexCursor +2

Explain why two scientific papers disagree, every claim grounded in a verbatim source quote. Runs locally from the studydiff-mcp npm package. Needs 2 environment variables to run.

not rated 1 17d ago A tokens not measured original MIT

patchistry-mcp-server

739

patchistry/patchistry-mcp-server

MCP server Claude CodeCodexCursor +2

DTC modular hat brand MCP server. 6 tools for AI agents: catalog, curated builds, shipping. Runs locally from the patchistry-mcp-server npm package.

not rated 1 3mo ago A tokens not measured original MIT

mcp-nevent

740

nevent-dev/mcp-nevent

MCP server Claude CodeCodexCursor +2

Talk to your live-events CRM (campaigns, analytics, paid ads, segments) in Claude and ChatGPT. Runs locally from the mcp-nevent npm package. Needs 2 environment variables to run.

not rated 1 3d ago A tokens not measured original MIT

querypilot

741

nickklos10/QueryPilot

MCP server Claude CodeCodexCursor +2

Eval-driven SQL reliability for AI agents. Runs locally from the querypilot Python package. Needs 6 environment variables to run.

not rated 1 2mo ago A tokens not measured original MIT

claude-bridge-mcp

742

rahuljava2807/claude-code-bridge-mcp

MCP server Claude CodeCodexCursor +2

Pipes Claude Code session output into Claude Chat — no more copy-paste between tools. Runs locally from the @rahulonramp2807/claude-bridge-mcp npm package.

not rated 1 5mo ago A tokens not measured original MIT

mcp-openapi-runner

743

saurav61091/mcp-openapi

MCP server Claude CodeCodexCursor +2

Turn any OpenAPI 3.x spec into callable MCP tools for Claude. No custom integration code needed. Runs locally from the mcp-openapi-runner npm package.

not rated 1 5mo ago A tokens not measured original MIT

rancher-mcp

744

rex/mcp-rancher

MCP server Claude CodeCodexCursor +2

Capability-aware MCP server for Rancher-managed Kubernetes: discovery, generic resource access, and curated operator workflows across the Norman and Steve planes (Rancher 2.6.5–2.9.3). Runs locally from the rancher-mcp Python package. Needs 4 environment variables to run.

not rated 1 6d ago A tokens not measured original MIT

At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: