Kargo automates promotion across dev, staging, and prod with approval gates and verification. Open source, built by the team behind Argo CD. Download now. Learn more →
Top 23 TypeScript llm-tool Projects
-
OmniRoute
Never stop coding. Free AI gateway: one endpoint, 160+ providers, RTK+Caveman stacked compression up to ~95% eligible context savings, smart auto-fallback, MCP/A2A, multimodal APIs, Desktop/PWA.
Project mention: Test OmniRoute Fallbacks for Semantic Consistency, Not Just Availability | dev.to | 2026-07-21diegosouzapw/OmniRoute is an MIT-licensed AI gateway that advertises one endpoint across many providers and models, with quota-aware fallback and token compression.
-
AppSignal
AppSignal knows why the f*#k it crashed. Stop vibe-debugging. Every exception, every backtrace, grouped so you see patterns, not noise.
-
Project mention: EvanFlow – A TDD driven feedback loop for Claude Code | news.ycombinator.com | 2026-04-26
If you’re just looking for the TDD part - https://github.com/nizos/tdd-guard - is the only project I’ve come across that actually enforces it with hooks and blocks edits rather than relying on a prompt that gets context rotted away.
-
llm-space
A desktop app to prototype agent ideas, inspect every harness step, replay failures, and evaluate performance, all in one place. Local-first, cloud-ready for managed agents.
llm-space on GitHub
-
Project mention: Show HN: Agent harness that turns errors into shared genes | news.ycombinator.com | 2026-04-21
-
A quick web search brought me to this: a Claude skill for all the ISO 24495 “writing styles”... https://github.com/GaZmagik/iso-24495
-
nse-bse-indian-stock-market-data-mcp
NSE BSE Indian Stock Market Data MCP server — search, screen & analyze all 8,200+ NSE/BSE stocks with 52 tools: live quotes, financials, technicals, 326-ratio screener, FII/DII flows, options IV, IPOs & portfolio tracking. Works with Claude, ChatGPT, Cursor & any MCP client
Project mention: I built an open-source MCP server that gives any AI assistant live NSE + BSE stock data | dev.to | 2026-06-19🔗 https://github.com/Tapetide-hq/nse-bse-indian-stock-market-data-mcp
-
don-cheli-sdd
Don Cheli — SDD Framework. The most comprehensive Specification-Driven Development framework for AI agents. 88+ commands, 51 skills, 15 reasoning models. TDD mandatory, OWASP audit, Autonomous Mode, Crash Recovery, PRD Generator. Works with Claude Code, Gemini/Antigravity, Cursor, Codex, Warp, Amp, OpenCode, Continue.dev. ES/EN/PT.
Project mention: Don Cheli – AI writes tests before seeing the code (TDD as iron law) | news.ycombinator.com | 2026-03-28 -
Kargo
Stop Scripting Promotions. Start Shipping with Kargo. Kargo automates promotion across dev, staging, and prod with approval gates and verification. Open source, built by the team behind Argo CD. Download now.
-
cloakbrowser-mcp
⚡ CloakBrowser MCP server for AI agents: Playwright-powered browsing, clean tool forwarding, Docker support, and multi-session HTTP transport.
Project mention: CloakBrowser MCP: Playwright MCP tools with a CloakBrowser Chromium runtime | dev.to | 2026-06-06GitHub: https://github.com/swimmwatch/cloakbrowser-mcp
-
Project mention: Shepherd's Dog: A Game by the Most Dangerous AI Model | news.ycombinator.com | 2026-06-13
I've been working on a project that's over 150k loc of Rust at this point. https://dirge-code.github.io
You absolutely can have the LLM write maintainable code. A few tricks I use are to ask it to plan out features in phases, and then do a branch and a PR for each focused piece of work. It makes it a lot easier to review and understand what's happening.
I also ended up making a tool which lets the LLM get a high level perspective of the codebase, and then see parts that are structurally gnarly. I've been using it to do refactors and clean things up periodically. It helped a lot with keeping the architecture clean.
https://github.com/yogthos/wavescope-mcp
Adding features and evolving the codebase has not been a problem even at this scale.
-
showsignature
A CLI that extracts the useful structure from source files: signatures, imports, types, variables, comments, and Markdown sections.
-
openhop
Interactive data-flow diagrams your AI agent can write. One SKILL.md — install in Claude Code, Cursor, Codex, and 11+ other clients via `npx openskills install naorsabag/openhop`.
-
june-2025-coding-agent-report
Comprehensive evaluation of 15 AI coding agents (Cursor, Copilot, Claude, Replit, v0, Warp, etc.) with implementations, screenshots, and professional scoring. Published on Turing Post.
-
anythingmcp
Self-hosted MCP gateway & API-to-MCP bridge. Convert REST, SOAP, WSDL, GraphQL, SQL, APIs into MCP servers for Claude, ChatGPT, Gemini, Copilot, Cursor. OpenAPI, Postman, WSDL import, OAuth2, RBAC, audit log. 2Pre-built adapters (DHL, DATEV, Weclapp, Personio, Handelsregister ...). Source-available.
Project mention: Show HN: AnythingMCP – source-available gateway turning REST/SOAP/SQL into MCP | news.ycombinator.com | 2026-05-12- Self-hosted via `docker compose up`, or one-click on Railway / DO Marketplace. There's also a managed cloud if you'd rather not run it.
A few things I expect HN to push on, so let me address them up front:
1. License is BSL-1.1, not MIT/Apache. We're a small team and the AWS-strip-mining trap is real for infrastructure projects. BSL converts to Apache 2.0 in 2030, and you can self-host today for any internal use, personal use, dev, testing, evaluation. The only thing you can't do is resell it as a hosted SaaS. Sentry, MariaDB, CockroachDB, HashiCorp Terraform are on the same model.
2. "Yet another MCP server?" — almost the opposite. AnythingMCP is a gateway that fronts your existing systems and other MCP servers behind one endpoint. The closest neighbours are mcp-proxy and Composio, but with the SOAP/database/RBAC/audit angle.
3. SOAP support is real, not a wrapper. We parse the WSDL and generate one MCP tool per operation. We needed it for ERP integrations and it's the single feature that's saved us the most time.
Happy to answer anything. Docs: https://anythingmcp.com
-
Project mention: How HN: Claudestat – real-time Claude Code monitor with loop detection | news.ycombinator.com | 2026-06-16
-
Yggdrasil
Say it once. Rules your AI coding agent still follows three sessions later, checked before it moves on. Free local checks, keyless CI. (by krzysztofdudek)
Project mention: Show HN: Gives your AI agents a shared, searchable, persistent memory – locally | news.ycombinator.com | 2026-03-30 -
Download: reticle.run Source: github.com/fwdai/reticle I’d genuinely love feedback — what’s missing, what’s wrong, what you’d prioritize. And I’m curious: how are you building and testing your agents today? What does your workflow look like? Drop it in the comments — I’m always looking to learn how others are approaching this.
-
Project mention: The Future of API Bills: When Machines Pay Machines with x402 Protocol | dev.to | 2026-07-21
The x402 protocol support in WAIaaS is one piece of a larger picture around autonomous agent infrastructure — if you want to understand how the policy engine works in depth or how DeFi actions fit into an agent's capabilities, the GitHub repository has the full source and documentation. The OpenAPI reference at /reference is also worth exploring once you have the daemon running — it's interactive and covers all 39 API route modules.
-
Project mention: Show HN: Mimirs – persistent local memory for AI coding agents (MCP) | news.ycombinator.com | 2026-06-11
-
ToolRAG
Unlimited LLM tools, zero context penalties — ToolRAG serves exactly the LLM tools your user-query demands.
-
Ornn
Your all-in-one skills-as-a-service platform to manage your skills, auto-generate skills and use agent skills as simple as calling an API.
Project mention: Agents aren't users — why we built a skill life-cycle API instead of a marketplace | dev.to | 2026-07-05We open-sourced Ornn at github.com/ChronoAIProject/Ornn — Apache 2.0, TypeScript, HTTP + MCP, model-agnostic (Claude / GPT / Gemini / your own runtime). The above is the design rationale; the README has the quickstart and the comparison table against MCP servers / Smithery / npm.
-
bhived-mcp
bhived is an MCP server that gives AI agents shared memory, skills, and tool discovery. install once, works in Claude Code, Cursor, and 15+ other agents.
-
openrouter-skill
OpenRouter skill for AI agents: model discovery, multimodal chat, tool calling, routing, and starter templates.
Project mention: Show HN: OpenRouter Skill – Reusable integration for AI agents using OpenRouter | news.ycombinator.com | 2026-03-05 -
oraclaw
Deterministic decision-intelligence MCP server for AI agents — 17 tools (bandits/LinUCB, HiGHS LP/MIP, PageRank, Monte Carlo, CMA-ES, conformal). Sub-25ms. Zero LLM cost. 11 free, no key. Listed on the MCP Registry & Glama.
Project mention: Traces show what your agent did - a decision ledger shows what it was allowed to do | dev.to | 2026-06-25(That re-checkable-certificate idea is what I've been building into OraClaw — deterministic decision tools that return verifiable results — but the three-layer ledger above is framework-agnostic; it's worth wiring into whatever runtime you're on.)
-
SaaSHub
SaaSHub - Software Alternatives and Reviews. SaaSHub helps you find the best software and product alternatives
NOTE:
The open source projects on this list are ordered by number of github stars.
The number of mentions indicates repo mentiontions in the last 12 Months or
since we started tracking (Dec 2020).
TypeScript llm-tools discussion
TypeScript llm-tools related posts
-
Every Check Said Our GA4 Analytics Worked. It Had Never Recorded a Visit
-
Test OmniRoute Fallbacks for Semantic Consistency, Not Just Availability
-
We pre-registered, ran, and verified the macro ablation: information per joule, measured
-
Production-Ready MCP Servers in 60 Seconds (Auth, Rate Limits, Audit Logs Included)
-
Beyond RAG: Why Knowledge Engineering Becomes the Real Moat in the Agent Era
-
Claude Advanced Tool Use
-
A note from our sponsor - Kargo
akuity.io | 8 Sep 2026
Index
What are some of the best open-source llm-tool projects in TypeScript? This list will help you:
| # | Project | Stars |
|---|---|---|
| 1 | OmniRoute | 4,001 |
| 2 | tdd-guard | 2,327 |
| 3 | llm-space | 1,806 |
| 4 | PrismerCloud | 1,410 |
| 5 | iso-24495 | 169 |
| 6 | nse-bse-indian-stock-market-data-mcp | 64 |
| 7 | don-cheli-sdd | 57 |
| 8 | cloakbrowser-mcp | 56 |
| 9 | wavescope-mcp | 45 |
| 10 | showsignature | 44 |
| 11 | openhop | 42 |
| 12 | june-2025-coding-agent-report | 40 |
| 13 | anythingmcp | 39 |
| 14 | claudestat | 34 |
| 15 | Yggdrasil | 33 |
| 16 | reticle | 30 |
| 17 | WAIaaS | 30 |
| 18 | mimirs | 29 |
| 19 | ToolRAG | 28 |
| 20 | Ornn | 20 |
| 21 | bhived-mcp | 14 |
| 22 | openrouter-skill | 14 |
| 23 | oraclaw | 13 |
Sponsored
AppSignal knows why the f*#k it crashed.
Stop vibe-debugging. Every exception, every backtrace, grouped so you see patterns, not noise.
www.appsignal.com