frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Show HN: Ratel, give agents unlimited tools and skills without context bloat

https://github.com/ratel-ai/ratel
17•jack1689•2h ago
Hi HN! We're Giacomo and Roberto, authors of Ratel (https://github.com/ratel-ai/ratel)

We used to help SaaS companies build agents on top of their products. Whenever we wanted to expand the agents’ complexity/scope, by adding more and more tools and instructions, we always run in the same issue: context bloat, with frequent hallucinations and sky high token bills. So we started constantly engineering the agents, dynamically loading tools, splitting them into subagents, inventing our own way to support skills

And that's exactly when we started building Ratel: a library to let your agent keep its full catalog of tools and skills, but progressively disclosing only the few that actually matter for each turn. Now you can grow your agent's capabilities without breaking it or taking out a loan for it

People are already using it in production, with a user cutting their token cost up to 81% in the first month without compromising the accuracy

We support both keyword and semantic retrieval, all in-process and without any additional infra. Open source, framework-agnostic, exposes OpenTelemetry metrics, available for Typescript and Python

Benchmarks: https://benchmark.ratel.sh

Some cool things we did with this:

• One team's agent had up to 300+ tools dynamically loaded into context. Ratel cut their token cost 81% in month one. • Another team split into several subagents instead, one agent per task. It worked, until the swarm got slow and expensive. We fixed this with our skills.

We're both here all day. Tear it apart, especially if you're an AI or SWE running agents in production

Comments

bbg2401•1h ago
Anyone else suspect this is an entirely astroturfed comment section?
tortilla•1h ago
Yes. The comments read like positive ai vibes and their profiles seem empty and low karma.
vinci00•1h ago
Isn't this just RAG for tools? Also, MCP already has tool search, how is this useful in that case?
jack1689•1h ago
Hey, yes it's on the same family for sure! The difference is that it works with anything, not just MCPs, and it runs in-process without any additional infra (which usually isn't the case for other RAG solutions). Happy to hear your feedback if you try it out :)
tomhow•1h ago
[stub for offtopicness]
iustinai•2h ago
man this could save me so much money lol
jack1689•2h ago
Would love to hear your feedback if you can try it. We initially rolled out BM25 for tool search as it worked best for us internally. Recently rolled out also embeddings and an hybrid option that is being tested in production as we speak
KristianLentino•2h ago
Benchmarks looks very promising! I’ll try to test this new tool in the next few days thanks for sharing!
rstagi•2h ago
Thank you mate! Please share your feedback :)
atenareply•2h ago
This looks so nice!
atenareply•2h ago
I see you built the core in Rust with bindings to TS/Python. y?
jack1689•2h ago
Yes! I should have mentioned in the original post. It was actually built in Typescript at first, but then the performance were not good enough for production use cases. With Rust the footprint was way lower, and we managed tear the latency down from 200ms to 20ms for a single search
missmoss•1h ago
This is neat! It tackles a boring but real problem with agents. When you have too many tools, the model gets confused. This is a good search box for its tools instead of dumping everything into the prompt. Benchmarks is amazing!
jack1689•1h ago
Thank you! Let us know how it works if you try it out :) of course real world is another story, but we agree benchmarks are amazing indeed!
flaviobernoni•1h ago
really cool, first time I've seen anyone treat "which tools does my agent even get to see" as its own problem. Is the roadmap more about pushing deeper on the retrieval side over time, or mostly about covering more frameworks/languages?
jack1689•1h ago
Thank you! Right now we're adding some frameworks adapters to lower the adoption friction, and pushing deeper on new algorithms on the retrieval side. More languages support is not on the radar yet, but we accept contributions! :)

Show HN: Leaves – A text-UI disk usage treemap visualizer

https://github.com/patonw/leaves
32•patonw•3h ago•8 comments

Show HN: A modern port of Linux to a ten-year-old QWERTY phone

22•tmzt•2h ago•3 comments

Show HN: Be the ChatBOT

https://bethechatbot.com/
3•keito•22m ago•0 comments

Show HN: QBasic Gorillas (Repeeled)

https://gameswithtony.com/gorillas/
11•TonyAlicea10•4h ago•8 comments

Show HN: Galois connections for composable numeric casts in Rust

https://github.com/cmk/connections
7•partialsolve•2h ago•1 comments

Show HN: Tree, truth, druid, dryad, and tar share one Proto-Indo-European root

https://p.migdal.pl/tree-of-tree/
3•stared•1h ago•0 comments

Show HN: Algotrek – Algorithms visualized, problem shape shifting and a tutor

https://app.algotrek.io
9•nyggy•4h ago•2 comments

Show HN: Sentinel – open-source QA agent that reads your code before it clicks

https://blog.simbastack.com/announcing-sentinel/
12•asenna•3h ago•3 comments

Show HN: Wallflower – A Quiet Social Reader

https://thewallflower.app/about
6•tldrthelaw•4h ago•0 comments

Show HN: Firefox in WebAssembly

https://developer.puter.com/labs/firefox-wasm/
235•coolelectronics•22h ago•119 comments

Show HN: I built an open source alternative to Workflowy

https://dotflowy.com
2•campak•3h ago•0 comments

Show HN: Open-source AI app builder you can embed into your own SaaS

https://github.com/totalumlabs/ai-app-builder-open
5•francescjuille•3h ago•0 comments

Show HN: misa77 - a codec that decodes 2x faster than LZ4 (at better ratios)

https://github.com/welcome-to-the-sunny-side/misa77
154•nonadhocproblem•1d ago•47 comments

Show HN: PokeTokenBar – Raise a Pokémon with your AI token usage

https://github.com/chattymin/PokeTokenBar
2•chattymin•3h ago•1 comments

Show HN: Token Gobbler – the goofy game where you're the LLM

https://tokengobbler.vercel.app/
2•dschwede•3h ago•2 comments

Show HN: Tmux tab markers for Claude Code, Grok, and pi sessions

https://github.com/jhickner/tmux-agent-tabs
2•jhickner•3h ago•0 comments

Show HN: Docs.dev Your Own Hosted Docs Platform in Minutes

https://docs.dev
8•linktothenew•3h ago•4 comments

Show HN: A spectral live stream of NYC sunlight

https://nyc-stream.enody.lighting
4•carterpeterson•3h ago•0 comments

Show HN: One More Letter

https://playonemoreletter.com/
83•hmate9•19h ago•56 comments

Show HN: I Built an Accent Changer App

https://accentchanger.com/
5•wbemaker•7h ago•11 comments

Show HN: Ingestr CDC – open-source CDC replication in Go

https://bruin-data.github.io/ingestr/getting-started/cdc.html
2•karakanb•3h ago•0 comments

Show HN: Democr.ai – self-hosted agentic AI runtime with audit and RBAC

https://github.com/democr-ai/democrai
2•fabio2•3h ago•0 comments

Show HN: Trying to fix the web scraping industry's benchmark problem

https://github.com/usestring/web-data-frontier-benchmark
13•rohitshenoy•4h ago•0 comments

Show HN: Nous – give GTM agents one context graph across your tools

https://github.com/NousC/opennous
10•sethiaakash04•4h ago•9 comments

Show HN: I built a Clojure runtime on top of Chez Scheme

https://yogthos.net/posts/2026-07-02-jolt.html
3•yogthos•4h ago•0 comments

Show HN: Ramsford – AI butler for travel, cruise planning, and daily tasks

https://ramsford.ai/
2•ramsford1000•4h ago•0 comments

Show HN: I've built tldraw for HTML and Markdown pages (with encryption)

https://pastehex.dev/
2•pandrre•4h ago•0 comments

Show HN: I've built a words game based on binary search

https://hilogame.cc/
43•ludovicianul•5h ago•42 comments

Show HN: An MCP server that frames screen recordings and shots in Apple bezels

https://github.com/cwooddgr/bezelbub/tree/main/bezelbub-mcp
3•igneous_rocker•4h ago•2 comments

Show HN: Termaxa – I found my AI-agent gate had silently stopped gating Cursor

https://github.com/termaxa/termaxa
2•devdoc83•4h ago•1 comments