Footer

Your go-to destination for discovering trending open-source projects and uncovering the insights that matter

Legal

  • Terms and conditions
  • Privacy policy

More

  • Pricing
  • Advertise
  • onedump
  • gosnakego
  • Contact us

Social

  • X (Twitter)
  • GitHub
Trendshift owl
What they build speaks louder than words
© 2026 Trendshift, created by Julian Li with ❤️
SignalAdvertise
Trendshift
SignalAdvertise
Sign in
Trendshift owl

Discover trends that matter

SignalAdvertise
DailyWeeklyMonthlyYearly
Live mentionsTopics
RepositoriesDevelopers
InsightsStats

Busbar
Busbar
Bifrost
Bifrost
Omnigraph
Omnigraph
OfinTech Desk
OfinTech Desk
It's a plan
It's a plan

headroomlabs-ai/headroom

Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.

Visit GitHub
Python
80k
6.7k
95 contributors
Apache License 2.0
website
AchievementsView all →
FeaturedAdvertise to 180K+ monthly visitors
Busbar

Control what AI can do before it acts

Bifrost

Fastest Enterprise AI Gateway for Routing, Governing, and Securing all your AI Traffic

Omnigraph

Versioned graph database on object storage. Git-style workflows and multimodal retrieval

OfinTech Desk

An open-source AI workspace for finance and accounting teams, in which every figure is traced to its source

It's a plan

Open-source Linear alternative

An issue tracker where AI agents take the work you assign them

Momentum

How fast headroomlabs-ai/headroom is gaining stars, and where it lands on Trendshift and GitHub Trending

Activity and trending

Stars, forks and merged pull requests each day or month, lined up with the repository's rank on Trendshift Trending

Sep 19 – Nov 18, 2026

Stars +7,039 in 61 days

0200400600
514

Forks +817 in 61 days

02040
36

Merged PRs 977 in 61 days

050100
69

Trendshift Trendingwhere the repository ranked each day it made the top 25

Top 25 in Python

1122816
Sep 19Sep 27Oct 5Oct 13Oct 21Oct 29Nov 6Nov 14
The left and right arrow keys step from day to day; Home and End go to the first and the last

Stars, forks, merged pull requests and Trendshift Trending rank, by day
DayStars gainedForks gainedMerged PRsPython
Sat, Sep 19, 2026+142+121Not trending
Sun, Sep 20, 2026+162+150Not trending
Mon, Sep 21, 2026+154+2416Not trending
Tue, Sep 22, 2026+118+191Not trending
Wed, Sep 23, 2026+116+1212Not trending
Thu, Sep 24, 2026+78+1265Not trending
Fri, Sep 25, 2026+88+1238Not trending
Sat, Sep 26, 2026+69+92Not trending
Sun, Sep 27, 2026+75+130Not trending
Mon, Sep 28, 2026+72+148Not trending
Tue, Sep 29, 2026+73+1222Not trending
Wed, Sep 30, 2026+79+924Not trending
Thu, Oct 1, 2026+60+1259Not trending
Fri, Oct 2, 2026+50+916Not trending
Sat, Oct 3, 2026+63+1115Not trending
Sun, Oct 4, 2026+67+917Not trending
Mon, Oct 5, 2026+35+1135Not trending
Tue, Oct 6, 2026+67+614Not trending
Wed, Oct 7, 2026+69+1013Not trending
Thu, Oct 8, 2026+162+1439Not trending
Fri, Oct 9, 2026+86+1810Not trending
Sat, Oct 10, 2026+83+852Not trending
Sun, Oct 11, 2026+153+140Not trending
Mon, Oct 12, 2026+63+55Not trending
Tue, Oct 13, 2026+109+1316Not trending
Wed, Oct 14, 2026+122+131Not trending
Thu, Oct 15, 2026+77+1210Not trending
Fri, Oct 16, 2026+78+1323Not trending
Sat, Oct 17, 2026+62+62Not trending
Sun, Oct 18, 2026+311+330#11
Mon, Oct 19, 2026+323+260#22
Tue, Oct 20, 2026+514+360#8
Wed, Oct 21, 2026+499+368#16
Thu, Oct 22, 2026+318+300Not trending
Fri, Oct 23, 2026+101+150Not trending
Sat, Oct 24, 2026+75+751Not trending
Sun, Oct 25, 2026+83+90Not trending
Mon, Oct 26, 2026+116+222Not trending
Tue, Oct 27, 2026+113+120Not trending
Wed, Oct 28, 2026+114+1913Not trending
Thu, Oct 29, 2026+115+1734Not trending
Fri, Oct 30, 2026+111+1112Not trending
Sat, Oct 31, 2026+90+130Not trending
Sun, Nov 1, 2026+95+130Not trending
Mon, Nov 2, 2026+102+100Not trending
Tue, Nov 3, 2026+79+1421Not trending
Wed, Nov 4, 2026+109+1269Not trending
Thu, Nov 5, 2026+81+1156Not trending
Fri, Nov 6, 2026+83+531Not trending
Sat, Nov 7, 2026+59+107Not trending
Sun, Nov 8, 2026+83+1220Not trending
Mon, Nov 9, 2026+88+718Not trending
Tue, Nov 10, 2026+64+513Not trending
Wed, Nov 11, 2026+65+1118Not trending
Thu, Nov 12, 2026+107+123Not trending
Fri, Nov 13, 2026+109+104Not trending
Sat, Nov 14, 2026+76+1135Not trending
Sun, Nov 15, 2026+76+121Not trending
Mon, Nov 16, 2026+97+173Not trending
Tue, Nov 17, 2026+74+1242Not trending
Wed, Nov 18, 2026+77+00Not trending
Total+7,039+817977
  • 1#1 on the list
  • 7Rank that day
  • Not trending
1
  • GitHub trending history

    Where the repository ranked on GitHub Trending, one dot for each day it was on a list

    Jun 2 – Oct 21, 2026#1 on 3 days

    #1#5#10#15#20#25
    Top 10
    Jul 2026AugSepOctNov
    The left and right arrow keys step from day to day; Home and End go to the first and the last

    Mentions and achievements

    Where people are talking about headroomlabs-ai/headroom, and every badge it has earned
    1. Nov 161 post
      • 同样的答案,JSON 省 60-95% 的 token。 headroom 在内容进模型之前先压缩:工具输出、日志、文件、RAG 片段。库、代理、MCP server 三种接法。 7.9 万星。答案没变,账单变了。

        Read full on X, posted by @nanagayaxoxo

        @nanagayaxoxo · x.com

    2. Nov 121 post
      • Claude Code 用久了,额度大头常常花在重复的上下文上:同一段日志、同一个文件,一轮一轮反复发给模型。 GitHub 上最近有三个项目专门处理这件事,各管一段。按用法整理了一下👇

        Read full on X, posted by @xoyif63862452

        @xoyif63862452 · x.com

    3. Nov 111 post
      • 🚨 6 FREE GitHub repos that make AI 10x more useful 🤯 1. GBrain: a second brain for AI 2. LLM Wiki: self-updating notes 3. Social Research Skills: finds viral outliers 4. Cinematic Skill: code-built…

        Read full on X, posted by @bradlantic

        @bradlantic · x.com

    4. Nov 91 post
      • 4. MemPalace/mempalace → local-first memory. Stores past conversations word for word, finds them with semantic search. Setup: npx skills add MemPalace/mempalace 60,476 stars, MIT. Heads-up from its…

        Read full on X, posted by @Decoded08

        @Decoded08 · x.com

    5. Nov 61 post2 posts
      • 上下文一塞满,Agent 就开始瞎编,Token 账单还跟着飞? 想少烧上下文、把工具输出和长会话压实,先收藏这 6 个(不含已单点写过的 Mem0): 1 → Headroom(约 7.9 万星 · Apache) 工具输出 / 日志 / RAG 切片进模型前先压一刀,关键行尽量留住 2 → Context Mode(约 3.4 万星 · ELv2 源码可见) 本地插件:完整输出落本机…

        Read full on X, posted by @perry_weijie

        @perry_weijie · x.com

      • Agent 上下文老被工具输出撑爆?先压一轮再喂模型🧰 Headroom Labs 开源 headroom: 工具日志、文件、RAG 片段进 LLM 前先压缩——官方称 coding agent 大约少 20% token,压缩后是否漏信息,建议用自己的任务验证;库 / 代理 / MCP 都能接,约 7.9 万⭐ 仓库在这 #开源 #AI编程

        Read full on X, posted by @Coralaaaaa

        @Coralaaaaa · x.com

    6. 7 days earlier
    7. Oct 301 post
      • I found 3 free AI repos and the 2nd one saved me money 😭 Agent-Reach · Headroom · career-ops #aitools #github

        Read full on X, posted by @HAnetoh

        @HAnetoh · x.com

    8. 8 days earlier
    9. Oct 223 posts
      • 用 AI 写代码,token 账单吓到你了吗 GitHub 上 7.7 万 Star 的 headroom:本地先把日志和工具输出压缩一遍再喂模型,答案不变。 💻 一条命令套住 Claude Code、Cursor、Codex、Cline,不用改代码 • 代码 Agent 省 20% token,JSON 数据最多省 95% • 压缩全在本地跑,不上传代码,自带仪表盘看省了多少 仓库放一楼👇…

        Read full on X, posted by @mingzi206

        @mingzi206 · x.com

      • AI帮你查个报错,先把几千行日志读一遍,问题还没解决,额度先下去了 🫠 Headroom就是想省这部分 它会先压缩要交给AI的日志、文件和工具返回结果,再让模型读,原文留在本地,需要时还能取回来 🔎 比较适合接API、用AI编程工具的人,不是装上就能给所有聊天软件省额度 真要试,拿同一个任务对比一下花费和结果,钱省了,关键报错可别也省没了 😂 #AI工具 #Token #开源

        Read full on X, posted by @nmyx888

        @nmyx888 · x.com

      • Claude、Codex 的用量,能白涨多少? 装上这个免费工具,订阅费不变:Claude 的用量大概多出 70%,Codex 多出 25%;平时查日志、排故障多的,涨 1.3 倍不止。 工具叫 Headroom,GitHub 上 7.7 万星,开源免费。它做的事一句话能说清:AI 每次读的东西,先帮它压一遍再喂进去。 💰 具体省在哪(官方自测) · 排查故障、翻日志找问题:省 57% ·…

        Read full on X, posted by @cc6pazpo

        @cc6pazpo · x.com

    You might also like

    Repositories close to headroomlabs-ai/headroom
    FeaturedAdvertise to 180K+ monthly visitors
    It's a plan

    Open-source Linear alternative

    An issue tracker where AI agents take the work you assign them

    Omnigraph

    Versioned graph database on object storage. Git-style workflows and multimodal retrieval

    Busbar

    Control what AI can do before it acts

    Bifrost

    Fastest Enterprise AI Gateway for Routing, Governing, and Securing all your AI Traffic

    OfinTech Desk

    An open-source AI workspace for finance and accounting teams, in which every figure is traced to its source

    open-compress/claw-compactor

    Python

    14-stage Fusion Pipeline for LLM token compression — reversible compression, AST-aware code analysis, intelligent content routing. Zero LLM inference cost. MIT licensed.

    2k232

    dshakes/distil

    Python

    Compression with a quality contract — cache-aware, causally-pruned LLM context compression for agentic runtimes, certified non-inferior across 7 domains. Works with any SDK.

    213

    orchetron/secondwind

    Rust

    Losslessly compress the tool output your AI agent sends the model, prove every value survived, and report the exact tokens removed. Run it as a proxy, a library, or middleware.

    414

    Supercompress/Supercompress

    JavaScript

    Query-aware context compression for LLMs and coding agents. Cut ~65% of input tokens.

    14826

    Aimaghsoodi/foveance

    Python

    A lossless token-optimization codec for LLM agents — cut input tokens up to 82%, nothing dropped, every fact kept. Drop-in proxy + one-line Python API for Claude Code, Codex, Cursor, OpenAI, Anthropic & Ollama. pip install foveance

    223

    snchimata/tokenfold

    Rust

    Private, provider-neutral context compression for LLMs. Reversibly compress schemas and logs inside your perimeter with zero third-party calls.

    32141

    fkiene/llmtrim

    Rust

    Local proxy that compresses your LLM API requests so you pay less, with no change to the answers. Trims wasted tokens from prompts, history, tool output, and code before they're sent: -31% input / -74% output, measured live. Any provider, no extra model calls. Also an MCP server and embeddable library (Rust, Python, Ruby, Kotlin, Swift, JS/TS).

    27525

    KathanModh259/latent-gate

    Python

    MCP server that shrinks logs, JSON, docs and web pages before your AI reads them. Offline, deterministic, ~3ms, no API key. 66% fewer tokens on a mixed dev corpus, 90%+ on logs.

    252

    jia-gao/leanctx

    Python

    Drop-in prompt compression for production LLM apps. Cut your token bill 40-60% without changing your code. Python SDK, LLMLingua-2, MIT.

    3237