tsouth89/conduit - RepoRadar

Score8.5

Popularity35.0

Risknone

TierGold

Score breakdown

Usefulness9.0

Novelty9.0

Momentum7.0

Maturity7.4

Open-source/build8.4

Evidence8.0

Workflow potential10.0

Setup ease8.8

Popularity is tracked separately. Support, ads, sponsorships, and tips never affect these signals.

Why it matters

Useful for AI coding-agent power users (Claude Desktop / Claude Code, Cursor, VS Code / GitHub Copilot, Codex CLI / Codex app) who run multiple MCP servers and hit the context-bloat wall — eager registration of all N×M tools blows past the context window in seconds, and the agent wastes tokens re-reading tool descriptions it never uses; conduit sits between the client and every MCP server and lazy

Who should use it

AI coding-agent power users (Claude Desktop / Claude Code, Cursor, VS Code / GitHub Copilot, Codex CLI / Codex app) who run multiple MCP servers and hit the context-bloat wall — eager registration of all N×M tools blows past the context window in seconds, and the agent wastes tokens re-reading tool descriptions it never usesUsers who care about **credential security** — auth profiles are first-class, written to the **OS keychain** (macOS Keychain / Windows Credential Manager / Linux Secret Service via D-Bus) rather than plaintext, and the gateway forwards them to the underlying server at startUsers who need **per-server ACLs** — allow/deny rules are ACL-style regexes against tool names (e.g. `allow = ["^search", "^read_"]`, `deny = ["^delete_", "^write_secrets"]`) so a coding agent can be sandboxed to read-only on a per-server basis without dropping the server entirelyUsers who need **sampling / elicitation control** — the gateway decides per-server whether the underlying server is allowed to make sampling requests back into the host LLM (`sampling = "off" | "local" | "remote"`) and whether the gateway should surface server-initiated elicit prompts to the user (`elicitation = "off" | "auto-approve-low-risk" | "prompt"`)Security / observability teams — every `call_tool` request is appended to a structured audit log with server, agent request id, tool name, args (with secret-arg redaction), response status, and wall-clock durationUsers who care about **local trust boundary** — no cloud relay, no telemetry, no central auth service, the gateway runs as a long-lived local process (systemd / launchd / Windows Service)Users who want a live dashboard — React + Vite UI at `localhost:7575` for live tool search, recent calls, audit log review, and per-server config editsEvaluation: `git clone && cd conduit && ./scripts/setup.sh && conduit serve` then add the bundled stdio entry to Claude Desktop / Claude Code / Cursor / VS Code / Codex CLI / Codex app

Who should skip it

Skip tsouth89/conduit unless the captured evidence suggests it solves a problem you are actively working on.

About this signal

tsouth89/conduit is tracked by RepoRadar as a local mcp gateway that lazy-coll in the MIT local MCP gateway that sits between every AI section. It was first seen on 2026-06-25 and last updated on 2026-06-25. The current verdict is 'try now' with a Gold tier and easy setup difficulty. tsouth89/conduit leads on workflow potential (10.0) and practical usefulness (9.0); its lowest signal is momentum (7.0), so factor that in before investing setup time. This page summarizes the evidence RepoRadar has captured from captured source metadata. The score, tier, risk label, and verdict on this page are never influenced by sponsorship, ads, or tips — they reflect only the usefulness, popularity, novelty, momentum, maturity, and evidence signals described in the RepoRadar methodology.

How this item is evaluated

RepoRadar assigned tsouth89/conduit a composite score of 8.5 out of 10, placing it in the Gold tier. This score combines weighted sub-signals: usefulness (35%), novelty (18%), momentum (14%), maturity (10%), open-source/build quality (7%), evidence quality (6%), workflow potential (6%), and setup ease (4%). Popularity is tracked separately at 35.0 and never affects the composite score or tier. The risk label of 'none' reflects inherent user-impacting hazards, not generic novelty. Items with no risk flag may still require normal code review before production use.

Putting this into practice? Read How to vet an AI agent or MCP server before you wire it in for the checklist behind this score.

Risk explanation

No inherent user-impacting risk is flagged from the captured evidence.

Evidence links

github.com

Closest alternatives / related signals

conduittsouth89mcp-gatewaymcp-servermcp-routerlazy-discoverymeta-toolssearch-tools