tare MCP server
Commit Nov 9, 2026Reduces LLM context window usage by 44–73% through lossless, cache-safe, and output-aware compression for coding agents.
Also known as mstuart/tare
- Stars
- 5
- Tools
- —
- Downloads, 30 days
- —
- Latest version
- —
- Last commit
- Nov 9, 2026
- License
- MIT
What is tare MCP?
tare reduces LLM context window usage by 44–73% through lossless, cache-safe, and output-aware compression for coding agents.
Summary generated from the project's README
About tare MCP server
Developers building coding agents can reduce context window usage by 44% to 73% using tare. This tool sits between an agent and a model API to compress tool outputs, logs, and JSON into denser forms. It is designed to be lossless by default, meaning information is only dropped if you explicitly opt into methods like AST code skeletonization or telegraphic natural language.
Users can deploy this as a local proxy by pointing an agent's base URL to the tare-proxy, which supports Anthropic and OpenAI endpoints. It is also available as a CLI with subcommands like compress and skeletonize, or as an MCP server for clients like Claude Desktop and Cursor. The system is cache-aware, detecting breakpoints to ensure provider prefix caches remain effective, and it is output-aware, backing off compression if it detects a spike in model verbosity.
Integration is flexible through the tare-mcp server, which requires no API keys as it operates via stdio, or through Python and JavaScript libraries for in-process compression.
What you can do with it
- 1Compressing large source filesA developer can use skeletonization to strip function bodies while keeping the structural signatures of a file.
- 2Reducing log and JSON overheadYou can shrink large tabular data or log files to fit more information into the context window.
- 3Integrating with MCP clientsUsers can add compression and persistent memory to any MCP client without needing an API key.
Written from the project's README. Every use case is checked against a quote from it.
Configuration
TARE_UPSTREAM
| Name | Description | |
|---|---|---|
| TARE_UPSTREAM | — |
Automated checks
Ran Nov 14, 2026Full report →Not verified yet. A server needs a score of 80, no serious finding, and enough data for us to check. How checks work · What to fix for more points
- Source repository linked and reachablemstuart/tare is public and reachablegithub.com/mstuart/tare ↗Pass
- Package names this repository as its sourceNot checked
- Build provenance (Sigstore attestation)Not checked
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-11-09github.com/mstuart/tare ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNot checked
- No critical advisory in the package or its dependenciesNot checked
- No hidden instructions to the model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNot checked
- Remote endpoint answers the MCP handshakeNot checked
README
View on GitHubtare sits between your agent and the model API and shrinks the context window. Three properties make it different:
- Lossless by default — tool output, logs, and JSON are re-encoded into a denser equivalent form. Information is dropped only when you opt in (row-caps, telegraphic NL, AST code skeletonization).
- Cache-correct — provider prefix caches discount cached tokens ~10×, and one rewritten byte in the prefix forfeits it. tare detects the cache breakpoint and compresses only the dynamic suffix.
- Output-aware — over-compression makes models compensate with verbose output, so total cost can rise as input falls. tare watches output tokens per turn and backs off when they spike.
What you get
tare-proxy— point your agent's base URL at it; zero code changes. Speaks Anthropic (/v1/messages) and OpenAI (/v1/chat/completions).tareCLI —compress,skeletonize,compact-lossy, and 12 more subcommands (full reference).tare-mcp— compression + persistent cross-session memory as MCP tools, for any MCP client.- Libraries —
tare-core(Rust),tare-compress(Python),tare-ai(JS/TS).
Proof
All numbers are measured by running target/release/tare on the committed corpus with tiktoken
o200k_base (v0.13.0). Results live in crates/tare-bench/results/; reproduce with
python3 crates/tare-bench/run_proof.py.
| Name | Content type | Command | Input tokens | Output tokens | Reduction |
|---|---|---|---|---|---|
| cargo_packages | json_array | tare compact-lossy |
6,906 | 3,625 | 47.5% |
| ps_aux | tabular | tare compact-lossy |
1,545 | 802 | 48.1% |
| app_log | logs | tare compact-lossy |
13,217 | 6,551 | 50.4% |
| agent_context | agent_context | tare compress |
15,130 | 8,499 | 43.8% |
| server_rs | code | tare skeletonize --path server.rs |
5,930 | 1,582 | 73.3% |
| json_crush_rs | code | tare skeletonize --path json_crush.rs |
3,937 | 1,607 | 59.2% |
| readme_prose | prose | tare compact-lossy |
5,732 | 2,727 | 52.4% |
Code reads are ~67–76% of a coding agent's tokens (SWE-Pruner, ACL 2026),
so skeletonization is the biggest lever. Head-to-head harnesses against Headroom, LLMLingua-2,
lean-ctx, and RTK live in crates/tare-bench/benchmarks/.
Compared to
| Scope | Deploy | Local | Lossless default | Output-aware¹ | |
|---|---|---|---|---|---|
| tare | tools · logs · files · JSON · history | proxy · library · CLI · MCP | ✅ | ✅ | ✅ |
| Headroom | all context | proxy · lib · MCP | ✅ | ❌ (reversible via cache) | ❌ |
| RTK | CLI command outputs | CLI wrapper | ✅ | ❌ | ❌ |
| lean-ctx | CLI commands, MCP tools | CLI · MCP | ✅ | ❌ | ❌ |
| LLMLingua-2 | prose / RAG | library (ML model) | ✅ | ❌ | ❌ |
| OpenAI / Anthropic native compaction | conversation history | provider-native | ❌ | ❌ | ❌ |
¹ Output-aware = reads the model's output token count each turn and reduces compression aggression
when verbosity spikes (the x-tare-verbosity-spike signal in
crates/tare-proxy/src/server.rs).
When to use · When to skip
Great fit if you…
- run coding agents and want savings without losing information by default
- care about the provider cache staying warm (tare won't break the prefix)
- want code reads compressed structurally (signatures kept, bodies elidable on demand)
Skip it if you…
- only use a single provider's native compaction and don't need a cross-provider proxy
- run in a sandbox where a local proxy process can't run
Get started (60 seconds)
# 1 — install (no Rust toolchain needed)
curl -fsSL https://raw.githubusercontent.com/mstuart/tare/main/install.sh | sh # → ~/.local/bin
# or: npm install -g tare-ai
# or: docker pull ghcr.io/mstuart/tare
# or: cargo install tare-cli
# or: git clone https://github.com/mstuart/tare && cd tare && cargo build --release
# 2 — run as a proxy (point your agent's base URL at http://localhost:8787)
TARE_UPSTREAM=https://api.anthropic.com tare-proxy
# 3 — or use the CLI on any stdin
cat big.rs | tare skeletonize --path big.rs # drop fn bodies, keep structure
ps aux | tare compact-lossy --max-rows 30 --max-field 110
| Env var | Default | Meaning |
|---|---|---|
TARE_UPSTREAM |
https://api.anthropic.com |
upstream API base URL |
TARE_PORT |
8787 |
listen port |
TARE_RECENCY |
4 |
recent tool outputs always kept |
TARE_ENABLED |
true |
set 0/false for byte-exact passthrough |
TARE_CONTEXT_LIMIT |
200000 |
model context window (drives the fill dial) |
TARE_OUTPUT_HOLDOUT |
0 |
fraction of sessions left uncompressed (A/B for tare output-savings) |
TARE_ADMIN_TOKEN |
unset | secret required by admin endpoints and CLI admin commands; unset disables the admin API |
TARE_LOG |
unset | set it to log one line per turn with the compression report |
Response headers (x-tare-input-tokens, x-tare-net-tokens, x-tare-dropped, x-tare-aggression,
x-tare-verbosity-spike, x-tare-halted) report what each turn did; GET /admin/stats and
POST /admin/runtime-env expose live stats and hot config when TARE_ADMIN_TOKEN is set; clients
must send the token in x-tare-admin-token. Details: getting started.
Integrations
# pip install tare-compress
import tare
out = tare.compress(blocks_json, task="fix the login bug") # in-process, no proxy needed
// npm install tare-ai
import { withTare } from "tare-ai";
const client = new Anthropic(withTare({ apiKey: "..." })); // routes through the local proxy
Python ships all core transforms in-process plus adapters for litellm, ASGI, LangChain, Agno, and Strands; JS/TS ships proxy helpers for the Anthropic/OpenAI SDKs and the Vercel AI SDK. Full adapter reference: docs/integrations.md.
Use with your Claude subscription — MCP, no API key
The proxy forwards whatever auth the client sends — a billable x-api-key, or your Claude Pro/Max
subscription OAuth token when you point Claude Code's ANTHROPIC_BASE_URL at it
(scripts/live-smoke-sub.sh runs exactly this round-trip). Prefer no base-URL change at all? Run
tare-mcp: a local stdio server your agent calls as tools — it never calls the model itself, so it
needs no API key.
# easiest — no build; npx fetches the prebuilt binary on first run:
claude mcp add tare -s user -- npx -y -p tare-ai tare-mcp
{
"mcpServers": {
"tare": { "command": "npx", "args": ["-y", "-p", "tare-ai", "tare-mcp"] }
}
}
Paste the same block into any MCP client (Cursor, Codex, Claude Desktop's
claude_desktop_config.json, …). It exposes 10 tools: compression (tare_compress,
tare_skeletonize, tare_compact_lossy, tare_deref_images), a reversible tare_expand,
tare_stats, and cross-session memory — full list.
Wrap your agent
One command starts the proxy and launches your CLI agent through it — ENV-based and ephemeral:
tare wrap claude # start proxy + launch Claude Code through it
tare wrap codex --port 9000 # 12 agents supported: claude, codex, aider, goose, …
tare wrap claude --print # dry-run: show what would run
Full agent matrix and modes: docs/cli.md.
Status & limitations
- Tested — verified end-to-end against the live Anthropic API (a full proxy round-trip on a
Claude subscription, plus the MCP server over real stdio JSON-RPC), on top of 228 unit,
integration, and property tests;
fmt/clippy -D warnings/cargo denygate every commit. - Deploy it as a local sidecar — tare runs next to your agent and forwards your credentials
upstream without logging or persisting them; treat it as a trusted component on your own machine,
not shared multi-tenant infrastructure (SECURITY.md). Startup failures exit with a
clear
[tare-proxy] fatal: …message, never a panic backtrace.
Known edges
- Proxy and CLI token counts are approximate (
tare-tokenize, chars/4); the benchmark numbers above are measured with tiktoken. - The context-fill signal counts the serialized request (incl. JSON envelope), so it slightly over-estimates fill — conservative, errs toward compressing sooner.
- A
>2 MBstreaming response whose final usage event straddles the 64 KB tail buffer may skip one verbosity sample (non-fatal).
Non-goals — no trained ML text-compressor (no weights to download, no inference latency) and no
audio: transcribe externally and feed the transcript through tare compress.
Architecture
Nine crates: engine (tare-core), proxy + closed-loop controller (tare-proxy), CLI, MCP server,
SQLite memory, tokenizer, cache models, Python bindings, bench harness. Optional neural-embed
feature swaps keyword relevance for exact-cosine neural embeddings.
Diagram and crate guide: docs/architecture.md.
Contributing
See CONTRIBUTING.md. CI runs cargo fmt --check, cargo clippy -D warnings,
cargo test, and a release build — please make sure those pass locally.
License
MIT — see LICENSE.
tare MCP questions
How does tare handle context compression?
It uses lossless-by-default compression that re-encodes tool outputs, logs, and JSON into denser equivalent forms, while remaining cache-correct for provider prefix caches.
Can I use tare with my Claude subscription without an API key?
Yes, by running tare-mcp as a local stdio server, your agent calls it as tools without the server needing to call the model itself, thus requiring no API key.
Does tare break provider prefix caching?
No, it is designed to be cache-correct by detecting the cache breakpoint and only compressing the dynamic suffix.