distil MCP server
Context optimization for LLM agents: tool registry, result masking, budgeting, compaction.
1 stars84 downloads/wk
Reviews
Write oneNobody has reviewed distil yet.
If you have run it, two minutes of your experience saves the next person an afternoon.
distil tools
No tool declarations could be read from the package source. They show once the server is installed.
Public scan report
scanner v0.1.10 · 2026-10-30 · same rubric, same numbers if you re-run it
- Code scan5 source files scanned15/25
- –Live reliabilityno gateway calls yet and no remote to proben/a
- –Tool poisoningtools not inspected (local package is not executed); not countedn/a
- Auth qualitylocal package, no credentials required12/15
- Maintenancelast push 21 days ago15/15
- Maintainer identityregistry namespace matches repository owner7/10
Findings (2)
- mediumnpm install lifecycle script present
install.script - mediumnpm install lifecycle script present
install.scriptpackage.json: …/distil-mcp.js" }, "scripts": { "postinstall": "node bin/postinstall.js", "test": …
How distil compares
Against every graded AI and agents MCP server, from our own scans and uptime checks. Updated after each scan run.
- Safety score
- 75/100
- higher than 45% of 4,460 AI and agents servers
- GitHub stars
- 1
- more than 50% of 2,711 AI and agents servers
- Downloads a week
- 84
- more than 60% of 1,538 AI and agents servers with a package
Install distil in Claude Code, Cursor or VS Code
Runs npx -y @munhq/distil on your machine. Read the scan report first; the gateway never runs local packages.
claude mcp add distil -- npx -y @munhq/distil
What the publisher says
From the distil repository's README, as published. We do not edit it. Read it on GitHub
claude mcp add distil -- npx -y @munhq/distilNo account, no API key, nothing to configure. The package is a small wrapper that fetches the binary for your platform and verifies it against the published checksums; install.sh and a prebuilt binary remain for anyone without Node.
Cache writes are 2% of your tokens and 28% of your bill
That is not a claim, it is a measurement: 13,814 real agent sessions, 410,742 assistant turns, re-run on 2026-08-29. Every unique token in them was billed 550 times, because a request resends the whole history.
Which means the obvious move — compress the history — is usually the wrong one. Editing history invalidates the cached prefix from the edit onwards and converts reads at 0.1x into writes at 1.25x or 2.0x. You cannot compress your way out of context cost. You can only decline to admit tokens.
Every tool in this space publishes a savings percentage measured on its own fixtures. What none publishes is the denominator: what share of a real session it is allowed to touch, and what that share costs once prompt-cache pricing is applied. distil measures both on transcripts an agent actually wrote.
What is prior art, and what is not. The cache arithmetic below is not a discovery. Anthropic's context editing docs state that clearing tool results invalidates the cached prefix, and ship clearatleast so a clear only fires when it is large enough to pay for that. The break-even rule is published too: on a 5-minute cache, cleared tokens times requests-before-the-next-clear must exceed 11.5 times the tokens you keep. The table in this README reproduces that rule exactly — it was derived independently, which is a check on the arithmetic, not a contribution.
The gap is empirical. Every source says to calibrate against your own workload, and none ships a way to do it or publishes what the values turn out to be. That is what this crate is for: measuring the numbers you need in order to choose clearatleast, or to decide not to clear at all.
The full measurement
Measured 2026-08-29 over 13,814 local Claude Code transcripts — 410,742 assistant turns, 212.7M tokens of unique text. Reproduce it on your own corpus with distil-bench ~/.claude/projects; a corpus grows, so the date matters more than the decimals.
Shortened. The full README is on GitHub.
Nothing above is checked by us. What we check is on the safety report.
distil: common questions
- Is distil MCP server safe?
- Mostly: it is graded B (75/100). Read the distil safety report
- How do I install distil?
- It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
- Does distil need an API key?
- Not as far as the registry entry and our scan can tell: no credentials are declared or required.
- Is distil maintained?
- The last commit was 26 days ago (2026-10-09). The latest release is v0.3.4.
- What can I use instead of distil?
- Servers from other publishers that do the same job: Image MCP server.
- How does distil compare with other AI and agents MCP servers?
- safety score 75/100 (higher than 45% of 4,460 AI and agents servers); github stars 1 (more than 50% of 2,711 AI and agents servers); downloads a week 84 (more than 60% of 1,538 AI and agents servers with a package). Full comparison.
Alternatives to distil
Same job from other publishers: the closest match first, then the best rated.
- ImageAI image generation and editing with prompt optimization and quality presetsnot reviewedEstablishedB
