gateway
Skill1 starsChanged 45 days agoMITView on GitHub
One MCP endpoint for many MCP servers, with profiles, policy, and budgets. Use when a project has several MCP servers, when servers must be grouped per context, when tool calls must be restricted (allow/deny, read-only), or when auditing which tools a setup exposes.
7 other projects use a version of this.
Adopted by
Most recently taken up by
- theopenco/llmgatewaytheir own version, first seen Nov 6, 2026
- simota/agent-skillstheir own version, first seen Oct 23, 2026
- kurtosis-tech/kurtosistheir own version, first seen Oct 8, 2026
- joelhooks/joelclawtheir own version, first seen Oct 8, 2026
- Tght1211/lan-proxy-gatewaytheir own version, first seen Oct 8, 2026
- Quad4-Software/skillstheir own version, first seen Oct 3, 2026
Other versions7 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 7
- theopenco/llmgateway★ 1.7ksame name
- kurtosis-tech/kurtosis★ 558same name
- Tght1211/lan-proxy-gateway★ 543same name
- simota/agent-skills★ 96same name
- joelhooks/joelclaw★ 65same name
- INERATE/atelier★ 2same name
- Quad4-Software/skills★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository9 other files it gives its agents.
- commitsmithSkill
- context-budgetSkill
- injection-guardSkill
- json-lensSkill
- log-tailSkill
- repo-mapSkill
- secret-scanSkill
- tokensaverSkill
- web-snapshotSkill
The file
We split this skill into the rules it gives an agent
3 rules. 0 are common, 0 are distinctive, and 3 are very specific to this project.
0
Common
Most of the 6 edited versions we read keep these.
0
Distinctive
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 40
Check gateway_status first when tools seem missing — a failed server explains itself.
- Very specific: line 41
Use profiles (serve --profile docs) to keep the active tool set small.
- Very specific: line 43
Use gateway_budget before enabling a heavy profile.
“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
MCP gateway
Aggregate many stdio MCP servers behind one endpoint. Tools appear namespaced as server__tool; profiles select which servers are active; policy rules and read-only mode restrict what can run; budgets show the schema cost.
Setup (required before use)
The gateway serves from a project-local config:
npx -y mcp-gateway init
npx -y mcp-gateway add filesystem -- npx -y @modelcontextprotocol/server-filesystem .
npx -y mcp-gateway add tokensaver -- npx -y tokensaver-mcp
npx -y mcp-gateway serveMCP tools exposed by the gateway
| Tool | Use it for |
server__tool | Every allowed upstream tool, namespaced by origin. |
gateway_status | Health: which servers connected, tool counts, policy summary, schema cost. |
gateway_policy | The active rules and which tools are blocked, with reasons. |
gateway_budget | Token cost of aggregated tool schemas, ranked. |
Workflow
- Check
gateway_statusfirst when tools seem missing — a failed server explains itself. - Use profiles (
serve --profile docs) to keep the active tool set small. - Set
readOnly: trueper profile for research sessions. - Use
gateway_budgetbefore enabling a heavy profile.
Tips
- Policy rules support globs (
delete_*) and are first-match-wins. - Audit logging can be enabled per config (
audit.enabled) and read back withmcp-gateway stats.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

