Skip to main content
Glama
ATR (Automation for Token Reduction)

ATR (Automation for Token Reduction)

ATR: Automation for Token Reduction

ATR makes your AI use fewer tokens when it works with tools. Fewer tokens means lower bills and fewer "usage limit reached" messages.

It works with Claude Code, Codex, Claude Desktop, Cursor, Windsurf, claude.ai and ChatGPT. It's free. Run it on your own computer, or add the hosted connector to a chatbot in a minute, with no install: https://atr-mcp.vercel.app (connector link: https://atr-mcp.vercel.app/mcp).

ATR in a chatbot: GitHub issues, web search, PDFs, places and weather, with 30 issues cut from 53,724 tokens to 419 and a PDF from 9,962 to 263

The problem

When an AI uses tools (GitHub, a browser, a database, your test runner), two things waste tokens:

  1. Every tool's instructions are sent on every message, even tools the AI never uses. Ten MCP servers can mean 10,000+ tokens before you've typed anything.

  2. Tool results are huge. A list of GitHub issues, a web page or a test run can be tens of thousands of tokens, and all of it goes into the chat.

Related MCP server: Refract

What ATR does

ATR sits between your AI app and your tools:

  • It hides the tool list. The AI sees a few small ATR tools and searches for the tool it needs when it needs it.

  • It shrinks results. Before the AI sees a result, ATR removes the noise and keeps what matters. A failing test run keeps the failures and drops the passing tests. A web page keeps its text and links and drops the layout.

  • Nothing is lost. When ATR cuts something, it keeps the full result, and the AI can ask for any part of it.

 
 
 

Get started

You need Python 3.11 or newer.

pip install atr-mcp
atr setup

atr setup finds the AI apps on your computer, shows you what it will change, and asks first. Then it:

  1. moves your MCP servers behind ATR, and

  2. adds ATR to each app.

Restart your apps and you're done.

Command

What it does

atr setup --dry-run

Shows the plan without changing anything

atr setup --undo

Puts every app back exactly how it was

pip install --upgrade atr-mcp

Updates ATR

Every file atr setup changes also gets a backup copy (.atr-backup).

Just want to try it?

The demo connects a fake server with 50 tools (GitHub, Slack, logs, a database). Nothing leaves your computer.

claude mcp add atr -- atr serve --demo        # Claude Code
codex mcp add atr -- atr serve --demo         # Codex

Ask your AI to "list the open GitHub issues", then "show ATR stats" to see how many tokens it saved.


How much it saves

Live tests

We gave AI agents the same browser task with and without ATR: open Hacker News, click "new", then open a GitHub issues page. Each setup ran 3 times, and every answer was correct.

Agent

Without ATR

With ATR

Claude Code

$0.237 per task, 13 turns

$0.110 per task, 10 turns (54% cheaper)

Codex

84k tokens (33k at full price)

134k tokens (18k at full price)

Codex is mixed: with ATR it sends about half as many full-price tokens, but more tokens overall.

Single results

What the AI receives, in tokens, measured with python -m benchmarks.bench_atr:

Result

Without ATR

With ATR

Saved

1,000 GitHub issues

429,203

879

99.8%

2,000-line server log

61,072

208

99.7%

git diff with a lockfile change

22,294

207

99.1%

Web page in a browser (Hacker News)

12,023

4,046

66%

50 tools' instructions, on every message

12,056

1,508

87.5%

pytest -v, 303 tests, 3 failing

6,685

675

90%

pip install, 60 packages

5,977

697

88%

In each case, the parts that matter are kept: errors, failing tests, changed code lines, page text and links.

When ATR helps, and when it doesn't

  • Helps most: you use MCP servers that return big results, like GitHub, browsers (Playwright), databases, logs or web pages.

  • Helps a little: plain coding in Claude Code or Codex with no MCP servers. Those apps already trim their own commands, so in our test ATR made no difference there.

  • Can't help: your chat history and the AI's own replies. An MCP server only sees tool calls, never the conversation.


ATR's tools

The AI uses these. You don't need to call them yourself.

In the free hosted connector

Tool

What it does

atr_fetch

Reads live web pages, PDFs, Office documents and JSON as compact text

atr_search

Searches the web with DuckDuckGo, Bing and Brave at once and returns one short merged list (about 60 tokens a result); no API key needed

atr_places

Finds local businesses and places with free OpenStreetMap data; no API key needed

atr_site

Inspects a website or searches its public sitemap and navigation

atr_weather

Gets current weather and up to seven forecast days; no API key needed

atr_query

Retrieves more from a large result ATR kept

atr_stats

Shows how many tokens ATR has saved

For your MCP servers

Tool

What it does

atr_find_tools

Searches all your servers' tools and returns the few that fit the task

atr_read

Runs a tool that only reads, and shrinks the result

atr_call

Runs a tool that changes something (your app can ask you first)

atr_query

Reads more of a result that was cut, such as one page, one field or a search

For coding

Tool

Instead of

atr_repo_map

Opening files one by one to learn a project

atr_search_code

Pages of search results

atr_read_file

Reading a whole file to see one function. It also says "unchanged" if the AI reads the same part twice

atr_diff

A git diff full of lockfiles

atr_run

Thousands of lines of test or build output

atr_stats

Shows how many tokens ATR has saved


Safety

  • Secrets are never read, searched or shown, including .env files, private keys, cloud credentials, and files named like api_keys.json. This also applies to tools behind ATR and to atr_diff.

  • Reading and changing are separate. Tools that only read go through atr_read. Anything that changes something goes through atr_call, so your app can ask you before it runs. atr_run (shell commands) always asks.

  • Files stay in allowed folders. File tools only see the folders listed in your settings.


Using ATR in claude.ai or ChatGPT

No install: add the hosted connector https://atr-mcp.vercel.app/mcp the same way (Settings → Connectors → Add custom connector, no sign-in). Then do the one-time step below.

Tell your chatbot to use it. Chatbots pick their own built-in web tools first, and in tests Claude never chose ATR by itself. Add this once under Settings → Profile → personal preferences (or in a Project's instructions):

When I ask you to search the web or read a link, use atr_search and atr_fetch instead of the built-in web tools.

Or skip the setup and ask your question and add "use atr", like "What is the latest on the MCP spec? Use atr." (or "use atr_search", "use atr_fetch"). This is only needed for the web tools. Tools you put behind ATR (GitHub, for example) are found by name.

For your own tools behind ATR, you need a link of your own:

These apps can only reach servers on the internet, so ATR needs a public link.

pip install "atr-mcp[http]"
atr serve --http                               # prints a link with a secret in it
tailscale funnel --bg 8789                     # a permanent public https link (or: cloudflared tunnel --url http://127.0.0.1:8789)

Add https://<your-link>/mcp/<secret> as a connector:

  • claude.ai: Settings → Connectors → Add custom connector (no sign-in)

  • ChatGPT: Settings → Apps & Connectors → developer mode → create a connector (no sign-in)

Keep the link private. Anyone who has it can use your tools. File reading and shell commands are off over the internet unless you turn them on with --read-files and --allow-run. To change the secret, delete ~/.atr/http_token.


Settings

atr setup writes these for you. You can also edit ~/.atr/config.json yourself:

{
  "servers": {                          // the MCP servers ATR runs for you
    "github": { "command": "npx", "args": ["-y", "@modelcontextprotocol/server-github"] }
  },
  "budget": 1500,                       // most tokens one result may use
  "level": 2,                           // 1 = keep almost everything, 3 = shrink hard
  "pin": ["github__create_issue"],      // tools to always show (skips the search step)
  "read_only": ["github__list_*"],      // tools that only read, if a server doesn't say so itself
  "read_roots": ["~"]                   // folders the file tools may open
}

More

No MCP at all? atr compress shrinks a file or your clipboard before you paste it into any chat:

atr compress dump.json --copy
pbpaste | atr compress --copy

Set up by hand: atr init writes the settings file, and atr snippet prints the entry to add to your app's MCP config.

"No module named 'atr'" on macOS: run chflags nohidden .venv*/lib/python*/site-packages/*.pth, or use the entry from atr snippet.

Limits:

  • Online servers that need a sign-in (OAuth) can't go behind ATR yet, so atr setup leaves them in your app.

  • ATR lowers what each message costs. It doesn't change your plan's limits.


For developers

git clone https://github.com/JacobBhatt12/atr && cd atr
pip install -e ".[dev]"
python -m pytest atr/tests -c /dev/null       # 44 tests
python -m benchmarks.bench_atr                # the savings table above

File

What's in it

atr/server.py

The MCP server your app talks to

atr/remote.py

The same server over the internet (atr serve --http)

atr/downstream.py

Talks to the MCP servers behind ATR

atr/compress.py

Shrinks results

atr/stash.py

Keeps full results for atr_query

atr/router.py

Tool search

atr/code.py

Coding tools

atr/setup.py

atr setup

Next up: remembering repeated tool calls so the same result isn't sent twice, sign-in support for online servers, and a small local model that summarizes long text when cutting would lose meaning.

License: MIT

Available Tools

10 tools
atr_callC
Destructive

Run a tool that changes something ([w] in atr_find_tools). Result is compressed.

Inputs: tool (required), budget, arguments.

atr_diffA
Read-only

What changed, compactly: git diff with 1 line of context, lockfiles/generated/binary files listed but not shown. Default: uncommitted changes. Use instead of git diff.

Inputs: ref, file, path, budget, +1 more.

atr_find_toolsA
Read-only

Find tools on the connected servers for a task. Returns compact signatures (* = required).

Inputs: query (required), k, detail.

atr_queryB
Read-only

Read part of a stashed result by ref. Data: path/where/fields/group_by. Text: search (regex).

Inputs: ref (required), path, limit, where, +4 more.

atr_readB
Read-only

Run a read-only tool ([r] in atr_find_tools). Result is compressed; big ones are stashed.

Inputs: tool (required), budget, arguments.

atr_read_fileA
Read-only

Read part of a local file. Pass lines= (e.g. from a search hit) or search= to get only what you need; with just a path, big code files return an outline instead of the whole file.

Inputs: path (required), fresh, lines, budget, +2 more.

atr_repo_mapA
Read-only

Project layout with each code file's top-level functions/classes. Start here in a codebase.

Inputs: path, budget.

atr_runB
Destructive

Run a shell command (tests, build, lint) and get compressed output: failures kept, noise cut.

Inputs: command (required), cwd, budget, timeout.

atr_search_codeA
Read-only

Find where something is defined or used, before reading any file. Regex, grouped by file (path:line: text); context=3 shows the lines around each hit, often enough to answer.

Inputs: pattern (required), glob, path, budget, +1 more.

atr_statsB
Read-only

Tokens ATR has saved in this session.

Inputs: no named parameters.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.5/5.0

    Scored across 10 tools

    Disambiguation4/5

    Most tools have clearly distinct purposes: discovery, read-only execution, write execution, stashed-result querying, local file reading, code search, repo mapping, diffing, shell running, and stats. There is some potential confusion between atr_read and atr_read_file, and between atr_call and atr_run, but the descriptions mostly clarify the boundaries.

    Naming Consistency4/5

    All tools use a consistent atr_ prefix and snake_case naming, which makes the set easy to scan. The only minor deviation is that a few names like atr_repo_map and atr_stats are noun-oriented rather than strict verb_noun forms, but the overall convention is predictable.

    Tool Count5/5

    Ten tools is well within the appropriate range for this server. Each tool covers a distinct part of the token-reduction workflow, and none appear redundant or trivial.

    Completeness4/5

    The set covers tool discovery, read/write execution, stashed result querying, file reading, code search, repo mapping, diff inspection, shell execution, and savings stats, which is strong lifecycle coverage for the stated purpose. Minor gaps exist around explicit stash management or direct local file editing, though agents can likely work around these via shell or connected tools.

    Maintenance

    ActivityNo data
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      A
      maintenance
      A proxy server that wraps existing MCP servers to significantly reduce token consumption by compressing tool descriptions into a two-step interface. It enables users to integrate extensive toolsets without exceeding context limits or incurring high API costs.
      143
      Apache 2.0
    • A
      license
      A
      quality
      C
      maintenance
      MCP proxy that compresses tool schemas on the fly. Up to 98% token reduction, 100% signal preserved verified after every compression. Zero LLM calls, fully deterministic.
      5
      3
      MIT
    • A
      license
      Not graded
      quality
      C
      maintenance
      Intelligent MCP proxy server that reduces context bloat by serving only the tools your AI actually needs through semantic search and a fixed two-tool surface.
      34 PyPI
      14
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables MCP-compatible developer environments to shrink verbose tool outputs—test logs, git diffs, directory listings, and file dumps—before they reach the context window, preserving compiler errors, failure signatures, and decisive anchors. It runs either as direct callable tools or as a transparent proxy wrapping other MCP servers such as filesystem, git, or terminal providers.
      1
      MIT