- TypeScript 90.8%
- JavaScript 8.8%
- CSS 0.2%
- HTML 0.2%
| Filename | Latest commit message | Latest commit date |
|---|---|---|
07d6316fba* build(web): fail vite build on a browser-externalized Node built-in (#1769) Vite 8 (rolldown) externalizes a node:* / Node built-in that reaches the browser graph with a warning and ships a `module.exports = {}` stub — the build still succeeds, shipping a broken bundle. The runtime browser smoke (smoke:web:browser, #1615) only catches the subset where that stub is *called* at module init; a stub imported but never called ships silently and nothing gates it. Add a Vite build plugin that turns that specific browser-externalization warning into a hard `vite build` error, moving detection upstream into `npm run build` / `validate` and additionally catching the never-called case the runtime smoke can't see. - clients/web/server/browser-externalized-builtin-gate.ts: Vite-agnostic detection + error + per-build state machine (unit-tested to the ≥90 gate). - vite.config.ts: thin Vite wiring. A throw inside rolldown's onLog is swallowed (the one hook where it doesn't abort — verified against vite@8.0.0), so the gate records the warning in onLog and re-throws in buildEnd. Scoped to `vite build` (apply: 'build'); the Node runner build (tsup) is unaffected. - scripts/verify-build-gate.mjs + verify:build-gate (in npm run ci and the GitHub workflow): runs a real build with a node:fs probe and asserts it fails via the gate — the only check that catches the warning phrasing drifting in a future Vite bump and silently disabling the message-keyed gate. There is no stable log `code` on the rolldown warning (only message + plugin), so the gate keys off the documented message phrasing; the drift guard above is what keeps that safe across Vite versions. smoke:web:browser stays as the runtime backstop. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XNdjEPKLG637X8YmhDiEk5 * build(web): harden the browser-externalized-builtin gate per review (#1769) Review follow-ups on PR #1777: - Per-build state: add gate.reset() called from the plugin's buildStart, so a `vite build --watch` rebuild doesn't inherit a prior build's recorded warning (the plugin instance is reused across rebuilds). Makes the docblock true. - buildEnd(error) now only asserts on the success path — rolldown also calls buildEnd(error) when the build already failed for another reason, and throwing then would mask that real error with the #1769 message. - verify-build-gate.mjs: drop the shell `rm -rf` cache-clear entirely (verified the gate fires reliably without it — the earlier "caching" symptom was the onLog-swallow behavior, not a stale cache); add a result.error spawn-failure check so a missing `npx` isn't misdiagnosed as "not via the gate"; add SIGINT/SIGTERM handlers so an interrupt during the multi-minute build still restores the mutated entry; add a progress log before the spawn. - Match the double-quote style of the sibling server/ files. - Docs: list browser-externalized-builtin-gate.ts in the AGENTS.md server/ block and the clients/web/README server/ list (with why a build helper lives under server/), and add verify-build-gate to the root README scripts/ line. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XNdjEPKLG637X8YmhDiEk5 * build(web): round-2 review hardening for the #1769 build gate - Scope the gate to the browser (`client`) environment via applyToEnvironment, so a future SSR/node environment built from this config isn't failed for a legitimate node:* import — browser-only intent is now structural. - Report ALL externalized built-ins in one build, deduped, instead of only the first offender (N leaks no longer means N fix-and-rebuild cycles). browserExternalizedBuiltinError now takes string[]; gate uses a Set. - verify-build-gate.mjs: pin Vite with `npx --no-install` so the drift check can never silently run a registry-fetched Vite instead of the repo-pinned one (the whole point is version fidelity); append the probe instead of prepending (ES imports hoist, and appending won't demote a future leading directive); clarify the signal-handler comment (handlers are a backstop — a blocked spawnSync tears down via the process group first). - AGENTS.md: update both `ci` pipeline enumerations to include verify:build-gate and drop the now-false "Storybook is the only CI step left out" claim. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XNdjEPKLG637X8YmhDiEk5 * build(web): round-3 review polish for the #1769 build gate - Error message: name the transitive-dependency remedy (a resolve.alias to a browser shim) — "remove the import / gate behind the backend" doesn't apply when the node:* import lives in node_modules. - isBrowserExternalizedBuiltinLog is now a `message is string` type guard, so recordLog drops its redundant `!== undefined` clause while still typing the Set.add() — a real one-liner, not a coverage-only removal. - verify-build-gate.mjs: docblock now states WHY it builds the real config + entry (catches config-level regressions — plugin removed, onwarn suppression added — that a temp config would miss), so the minute-long build isn't later "optimized" into a temp entry and silently lose that coverage. Guard the entry read so a renamed main.tsx fails through fail() with an actionable message. On a botched restore, save the captured original to a sidecar .bak and point there instead of recommending `git checkout --`, which would discard uncommitted edits. - AGENTS.md: fix the muddled CI-steps sentence (GitHub CI runs the whole chain as separate steps, Storybook last) and note the applyToEnvironment client-only scoping in the gate prose. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XNdjEPKLG637X8YmhDiEk5 * build(web): round-4 review polish — sharper verify-build-gate diagnostics (#1769) All in scripts/verify-build-gate.mjs's failure paths: - Split the "build SUCCEEDED" failure into three diagnoses keyed off the captured output, so it names the right file: warning present but build passed → plugin not applying (removed from `plugins`, `applyToEnvironment` env-name mismatch, or an `onwarn` suppression); node:fs present but phrase absent → phrasing drift, update the constant; neither → probe never reached the graph. The old message hard-coded "phrasing drifted", which misdirects for the config-regression class this script exists to catch. Added applyToEnvironment to the docblock's config-regression list too. - Bound the build with a 10-min spawnSync timeout (+ SIGKILL); a timeout sets result.error (ETIMEDOUT) and reports via the existing branch, instead of burning to the GitHub job's 360-min default with no output. - Guard the probe write so a read-only/EACCES checkout fails through fail() with an actionable message instead of escaping as a raw stack. - SIGTERM now exits 143 (was 130, which is SIGINT's code). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XNdjEPKLG637X8YmhDiEk5 * build(web): round-5 review polish — self-consistent verify-build-gate (#1769) All in scripts/verify-build-gate.mjs: - Drift guard: assert the gate source file still contains the mirrored KNOWN_PHRASE literal, failing fast if the two drift. Without it, a stale KNOWN_PHRASE after a phrasing update would make the three-way diagnosis misreport a plugin-not-applying regression as a phrasing drift — the exact misdirection the diagnosis exists to avoid. - Make the .bak safety net reachable on the path that needs it most: restoreEntry now catches its own write failure and routes to a shared saveBackupAndFail helper, instead of letting the write escape the finally as a raw stack (which skipped the .bak fallback and left the probe injected in the entry). - Add the fourth "plugin isn't applying" cause to the diagnosis: a future Vite emitting the warning before the client env's buildStart reset clears it (same symptom as the other three). - Flag once, on the failing-check path, that clients/web/dist holds a probe build so a local debugger doesn't serve it unaware. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XNdjEPKLG637X8YmhDiEk5 * build(web): round-6 review polish — verify-build-gate literal contracts (#1769) All in scripts/verify-build-gate.mjs; one theme — the literals the script asserts on can each go stale independently of what they mirror: - Success key: assert on the gate's distinctive thrown-error prefix "Build failed (#1769)", not a bare "#1769". The issue number also appears in vite.config.ts comments, so a code frame from an unrelated build error there could otherwise make the check report OK with the gate dead — the one direction it must never get wrong (a false pass hiding a dead gate). - Drift guard now anchors on the BROWSER_EXTERNALIZED_BUILTIN_PHRASE assignment (regex) instead of a whole-file substring match, so the old wording lingering in a comment can't mask a changed constant; and its readFileSync of the gate module is guarded so a moved file fails through fail() actionably. - Extract PROBE_MODULE ("node:fs") used to build both the PROBE and the diagnosis check + messages, so changing the probe module can't silently misroute a phrasing-drift failure into "probe never reached the graph". - Comment the load-bearing diagnosis order (phrase before module-name). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XNdjEPKLG637X8YmhDiEk5 * build(web): round-7 review polish — drift-guard both mirrored literals (#1769) - Extract ERROR_PREFIX ("Build failed (#1769)") and drift-guard it against the gate module too, and use it as the success key. Previously only KNOWN_PHRASE was guarded; the prefix is mirrored across the same TS/.mjs boundary for the same reason, so rewording the gate's thrown error would make a correctly-firing gate report "the build broke for another reason" (benign direction, but the same misdirection the other guards close). - Write the restore-failure .bak under the OS temp dir, not next to the entry: a sidecar inside src/ is untracked and a `git add -A` (likely while recovering) would sweep it into a commit. - Branch-(b) remediation now says update BOTH the gate constant AND KNOWN_PHRASE, so a phrasing-drift fix lands in one pass instead of tripping the drift guard on the next run. - Note next to BROWSER_EXTERNALIZED_BUILTIN_PHRASE that it must stay a single string literal (the script's drift guard anchors a regex on the assignment). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XNdjEPKLG637X8YmhDiEk5 * build(web): round-8 review polish — safe recovery path + structural onLog order (#1769) - saveBackupAndFail no longer claims a backup it may not have written: it tracks whether the write succeeded and only points the user at the .bak in that case (otherwise it says the probe is still injected and to restore from VCS). Uses mkdtempSync for a unique dir instead of a fixed tmp filename, so a pre-existing/foreign-owned /tmp file can't make the recovery message point at someone else's content. - Guard the restore-verification readFileSync too, routing a failure through saveBackupAndFail so the whole restore path is uniformly safe (a raw throw there would skip the backup net). - vite.config.ts: enforce: 'pre' on the gate plugin so it runs ahead of the normal-plugin group in the onLog chain — a plugin whose onLog returns false filters the log for later plugins, so trailing the array would let a future log-filtering plugin silently blind the gate. Makes "the gate sees every log" structural rather than array-position-dependent. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XNdjEPKLG637X8YmhDiEk5 * build(web): round-9 review polish — prove the gate fired on the probe (#1769) The success key confirmed the #1769 gate fired, but not that OUR probe tripped it. If the repo already leaked a Node built-in, the gate would fire on that and the script would report OK even if the probe was tree-shaken or entryPath went stale — branch (c)'s failure silently inverted into a pass. Add an assertion that PROBE_MODULE appears among the gate's embedded offender warnings. Unreachable inside `npm run ci` (validate's build:web fails first on a pre-existing leak), but this script is a documented standalone command — run exactly when debugging a leak, the case where this mattered. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XNdjEPKLG637X8YmhDiEk5 * build(web): round-10 review polish — roll back a partial probe write (#1769) The probe writeFileSync uses the default 'w' flag (truncates at open), so a mid-write ENOSPC/EIO could leave clients/web/src/main.tsx empty or half-written — and its catch called fail() directly, the one path on this script that could destroy the entry rather than merely dirty it (every other read/write is already routed into the recovery net). Call restoreEntry() in that catch to roll the partial write back (falling through to the .bak net if the restore itself can't write), and correct the comment that claimed a write failure happens "before any mutation". Also widen the enforce:'pre' comment to note it orders ahead of Vite's core plugins too (harmless — the gate defines no resolve/load/transform). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XNdjEPKLG637X8YmhDiEk5 * build(web): round-11 review polish — don't let the rollback swallow the write error (#1769) Round 10's restoreEntry() in the probe-write catch swallowed the primary error: restoreEntry() on its own write failure calls saveBackupAndFail → fail → exit, so the "could not write the probe" diagnosis never ran when the rollback also failed — which is exactly the read-only-checkout / ENOSPC case (both writes fail at open). On a read-only checkout nothing was ever written, yet the developer was told the restore failed and pointed at a /tmp backup of an untouched file. Roll back inline instead: attempt the restore without exiting, track whether it succeeded, then always report the PRIMARY write error, appending "restore from version control" only when the rollback didn't take. Also pin the "Build failed (#1769)" error prefix as contiguous in a single string fragment (a comment symmetric with the phrase pin), since verify-build-gate.mjs drift-guards on it as ERROR_PREFIX. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XNdjEPKLG637X8YmhDiEk5 * build(web): round-12 review polish — don't let the pin comment defeat its own guard (#1769) The round-11 pin comment quoted "Build failed (#1769)" verbatim, so the prefix appeared twice in the gate module and verify-build-gate.mjs's whole-file includes(ERROR_PREFIX) drift guard passed on the comment alone — rewording the thrown error would have kept the guard green while a correctly-firing gate reported "the build broke for another reason." Reworded the comment to not repeat the full lead (matching how the phrase pin avoids repeating its phrase); the literal now occurs exactly once, restoring the guard's meaning. Also make the probe-write recovery message factual instead of inferred: read the entry back and only warn "restore from version control" when it actually differs from the original. On a read-only checkout both writes fail at open, so the file is untouched and the old hedge needlessly implied damage. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XNdjEPKLG637X8YmhDiEk5 --------- Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
| .claude | ||
| .github/workflows | ||
| clients | ||
| core | ||
| docs | ||
| scripts | ||
| specification | ||
| test-servers | ||
| .dockerignore | ||
| .gitignore | ||
| .npmignore | ||
| AGENTS.md | ||
| CLAUDE.md | ||
| Dockerfile | ||
| eslint.config.js | ||
| package-lock.json | ||
| package.json | ||
| README.md | ||
| tsconfig.base.json | ||
| vitest.shared.mts | ||
MCP Inspector
A developer tool for inspecting Model Context Protocol (MCP) servers. It ships as a single package, @modelcontextprotocol/inspector, that provides three ways to inspect a server:
- Web — a Vite + React + Mantine single-page app with a Node backend.
- CLI — a scriptable command-line client for automation, CI, and fast agent feedback loops.
- TUI — an interactive terminal UI built with Ink.
All three run through one global mcp-inspector binary:
npx @modelcontextprotocol/inspector # web UI (default)
npx @modelcontextprotocol/inspector --cli # CLI
npx @modelcontextprotocol/inspector --tui # TUI
Repo status. This is the v2 line of the Inspector (branch
v2/main). Themainbranch is the legacy v1 implementation (bug fixes only). v2 will eventually replacemain. SeeAGENTS.mdfor branch/board conventions.
Project layout
v2 is not an npm workspace. Each client under clients/* keeps its own package.json and node_modules; shared code lives in core/ and is consumed via a @inspector/core build-time alias (no package.json of its own). A single npm install at the root cascades installs into every client (see Setup).
inspector/
├── clients/
│ ├── web/ # Web client (Vite + React + Mantine). src/ = browser app; server/ = Node dev/prod backend
│ ├── cli/ # CLI client (tsup bundle, @inspector/core alias)
│ ├── tui/ # TUI client (Ink + React, tsup bundle)
│ └── launcher/ # Shared launcher — provides the `mcp-inspector` bin, dispatches to web/cli/tui
├── core/ # Shared code consumed via the `@inspector/core` alias (no package.json)
│ ├── auth/ # OAuth: providers, discovery, storage, mid-session recovery (browser/node/remote backends)
│ ├── json/ # JSON + parameter/argument conversion utilities
│ ├── logging/ # Silent pino logger singleton
│ ├── mcp/ # InspectorClient runtime, state stores, transports, config import
│ ├── react/ # React hooks over the state stores
│ └── storage/ # File I/O helpers for the OAuth persist backends
├── test-servers/ # Composable MCP test servers + fixtures used by integration tests
├── scripts/ # Root build/verify tooling (install cascade, smokes, verify-build-gate, pack:verify)
├── specification/ # Design/build specifications
├── AGENTS.md # Contribution rules for agents AND humans (see below)
└── README.md # You are here
Each client has its own README with client-specific detail: web · cli · tui · launcher.
Task-oriented guides live under docs/ — see Reviewing an MCP App, the CLI-first → one-shot-web recipe for automated App-tool review: --app-info probe → deep-link navigate → rendered widget, plus OAuth handoff and proxy support.
Setup
Requires Node >=22.19.0.
npm install # root install; postinstall cascades into every client
- Fresh clone: run
npm installat the repo root. - After a pull that changes a client's dependencies: re-run
npm installat the root to re-sync every client.
The cascade (scripts/install-clients.mjs) is dev-only — it exits early when the package is installed as a dependency, and the published tarball ships only each client's build/, so end users are unaffected. Set INSPECTOR_SKIP_CLIENT_INSTALL=1 to skip it.
Running during development
For day-to-day web iteration, run Vite directly from the web client (fast HMR, no launcher build needed):
cd clients/web && npm run dev
The launcher-driven scripts below run the built launcher, so build first (npm run build):
npm run web # prod web launcher against clients/web/dist
npm run web:dev # web launcher in --dev mode (Vite)
The @inspector/core shared package
core/ holds the logic shared by all three clients so that web, CLI, and TUI behave identically. Its entry point is the InspectorClient class (core/mcp/), which owns the connection to an MCP server, the request/response lifecycle, and a set of state stores; core/react/ exposes React hooks over those stores that both the web and TUI (Ink) React trees consume. OAuth (core/auth/) is factored into isomorphic logic plus browser/node/remote backends so the same flows work in the browser, in Node, and against a remote backend.
core/ intentionally has no package.json — it is not published on its own. Each client bundles it in via a @inspector/core alias:
- CLI / TUI:
esbuildOptions.aliasin theirtsup.config.tsmaps@inspector/core→ the repocore/directory, andnoExternal: [/^@inspector\/core/]inlines it into the bundle. - Web: the same alias in
clients/web/vite.config.tsfor the browser app and the Node backend runner.
Publishing core/ as its own package (e.g. for third parties to build on) is deliberately deferred — see issue #1636.
Web client: "dumb components" + Storybook
The v2 web client is built from presentational ("dumb") components — they accept data and callbacks as props and contain only display logic, with no direct data fetching or client state. State comes from the @inspector/core hooks, wired in near the top of the tree. This keeps components isolated, testable, and documentable.
That approach is what makes Storybook first-class here: every screen and element component has a *.stories.tsx file (96+ stories) that renders it against fixture props. Storybook play functions double as interaction tests, run headless in CI (npm run ci:storybook, Chromium via Playwright).
Styling follows a strict Mantine-first convention (theme variants and component props over CSS classes, --inspector-* CSS custom properties over raw color literals). The full rules live in AGENTS.md under React instructions — read them before touching web UI. Element components live in clients/web/src/components/elements/; theme variants in clients/web/src/theme/.
Test servers
test-servers/ provides composable MCP servers used by the integration and smoke suites, so tests exercise a real server over a real transport instead of mocks. A server is assembled from presets (fixture factories in test-servers/src/preset-registry.ts — tools, resources, prompts, tasks, elicitation, sampling, OAuth, …) and can be driven two ways:
- In-process — import the factories (
createTestServerHttp,createEchoTool, …) and run the server inside the test's event loop (used by the HTTP integration paths). - As a subprocess —
test-servers/build/test-server-stdio.jsis spawned as a real stdio child (used by the CLI smoke and stdio integration tests).
Configure a server declaratively with a JSON config (see test-servers/configs/*.json) selecting presets, then load it via --config. Because the servers are spawned as real subprocesses, the build output must exist first:
npm run test-servers:build # (from clients/web) → tsc -p test-servers, emits test-servers/build/
The Vite alias @modelcontextprotocol/inspector-test-server (in clients/web/vite.config.ts) points at test-servers/build/index.js so getTestMcpServerPath() resolves to a real .js path.
A streamable-HTTP server can also serve the modern (2026-07-28) protocol era via the SDK's createMcpHandler — set transport.modern in the JSON config (true for dual-era stateless serving, or { "legacy": "reject" } for modern-only strict), or pass modern on the ServerConfig for an in-process createTestServerHttp. This is what lets an Inspector connection negotiating protocolEra: "auto" | "modern" reach the modern leg (populated server/discover, sessionless). See test-servers/configs/modern-http.json. test-servers/configs/modern-mrtr-http.json additionally serves the mrtr_confirm tool (preset mrtr_confirm, createMrtrTool) over the modern leg: its handler returns inputRequired(...) embedding a form elicitation, so invoking it produces a real MRTR round-trip (input_required → the client fulfils the embedded elicitation and retries with a new id → complete). The Inspector drives MRTR manually (inputRequired: { autoFulfill: false }), so the embedded elicitation pauses at the pending-request modal (tagged "input_required") for you to answer, then the retry completes — useful for eyeballing both that pending-request UX and the Protocol view's MRTR conversation grouping. test-servers/configs/mrtr-showcase-http.json bundles every MRTR preset in one modern server for manual testing: mrtr_confirm (single round), mrtr_two_step (two elicitation rounds via requestState), mrtr_sample (embedded sampling → the Sampling panel), mrtr_roots (embedded roots/list, auto-answered silently from configured roots — no modal), mrtr_edge (an inputRequests-only round then a requestState-only round), and mrtr_loop (never completes → the MRTR_MAX_ROUNDS bound trips). (The legacy collect_elicitation preset calls server.elicitInput, which errors on the 2026-07-28 leg — server→client requests aren't allowed there; MRTR is the modern replacement.)
test-servers/configs/modern-network-http.json is the Network-tab showcase for the standardized HTTP headers and new error taxonomy (SEP-2243 / SEP-2575). It serves a get_weather tool whose city argument carries an x-mcp-header: "City" annotation (so a modern client mirrors it to Mcp-Param-City), plus four trigger_* tools that the modern leg's spec-error injector (transport.modern.injectSpecErrors: true) answers with a real HTTP status + JSON-RPC error body: trigger_header_mismatch → 400 / -32020, trigger_missing_capability → 400 / -32021, trigger_unsupported_version → 400 / -32022 (with data.supported), trigger_method_not_found → 404 / -32601. Connect to it with Protocol Era = Modern and open the Network tab to see the mirrored Mcp-* headers highlighted, sentinel values decoded, and each error rendered distinctly. Note: Mcp-Param-* mirroring is skipped by the SDK in the browser (detectProbeEnvironment() !== "browser"), so calling get_weather from the web client omits Mcp-Param-City and the strict server answers -32020 — the same tool is callable from the Node CLI/TUI, where mirroring is active.
test-servers/configs/xmcpheader-modern-http.json is the x-mcp-header Tools-tab showcase (#1632). It serves echo, a get_weather tool with a valid x-mcp-header: "City" annotation on its city argument, an invalid_header_tool whose annotation uses the header name "Bad Header" (a space makes it an invalid RFC 9110 token, so the whole tool definition is invalid), and a trigger_invalid_params tool that the modern leg's spec-error injector (transport.modern.injectSpecErrors: true) answers with a real -32602 Invalid params JSON-RPC error whose message is not about a missing tool. Connect with Protocol Era = Modern and open the Tools tab: get_weather's detail panel shows a "Mirrored request headers (SEP-2243)" section (city → Mcp-Param-City), and invalid_header_tool appears struck-through under an "Excluded (SEP-2243)" divider in the sidebar with the reason on hover (a conforming Streamable HTTP client MUST drop it from tools/list; the Inspector surfaces why). Under SDK v2 a tools/call that rejects with -32602 now renders as a distinct error panel rather than an isError result — headed "Unknown Tool" when the message names a missing tool (reproduce by calling a tool the server dropped from its list), or "Invalid Parameters" for any other -32602 (run trigger_invalid_params).
test-servers/configs/pagination-http.json is the page-by-page fetch showcase (#1721). It serves 12 tools, 12 resources, and 12 prompts (presets numbered_tools / numbered_resources / numbered_prompts, count: 12) with maxPageSize of 4 for each, so every list paginates into three pages. Turn on "Fetch Lists One Page at a Time" (Server Settings — the paginatedLists setting, or the Paginated switch in a list sidebar) and the Tools/Resources/Prompts lists load page 1 only (4 items) with a Load next page control and an N pages loaded status; each click fetches the next 4 and appends them, and Refresh resets to page 1. With the switch off (the default), the same lists auto-aggregate all three pages on connect.
test-servers/configs/advertised-extensions-http.json is the advertised-extensions showcase (#1739). It serves echo (always) and a get_weather tool that is gated on the io.modelcontextprotocol/tasks extension (extensionGatedTools): the tool is registered but starts disabled, and the server enables it on notifications/initialized only when the connected client declared that extension in its capabilities.extensions. Connect (the Inspector advertises the Tasks extension by default) and the Tools list shows both echo and get_weather. Open Server Settings → Advertised Extensions, uncheck Tasks (io.modelcontextprotocol/tasks), and reconnect: the client now advertises no extensions, the server never enables get_weather, and the Tools list shows only echo. This demonstrates the debugging knob — a server legitimately changing tool registration on what the client advertises. (Legacy stateful leg only; the modern per-request leg has no persistent oninitialized.)
test-servers/configs/logging-legacy-http.json and test-servers/configs/logging-modern-http.json are the logging era-fork showcase (#1629). Both serve logging: true plus a send_notification tool that emits a notifications/message at a chosen level; the legacy one is a plain streamable-HTTP server (logging/setLevel era) and the modern one sets transport.modern: true. Connect to the legacy server and open the Logs tab to get the session-scoped Set Active Level selector + Set button; calling send_notification streams the log into the panel. Connect to the modern one with Protocol Era = Modern and the same tab instead shows the Log Level per Request control — pick a level to opt in and the client stamps _meta["io.modelcontextprotocol/logLevel"] on every subsequent request (verify in the Network tab's request body); calling send_notification then streams the log into the panel over the request's SSE response. Set the control back to Off and the same call is silently gated — the request omits the logLevel key, so the log never arrives. That gating is faithful to the spec ("a server MUST NOT emit notifications/message for a request that didn't opt in") because send_notification emits through the SDK's request-scoped, threshold-aware extra.log (ctx.mcpReq.log): on the modern leg it reads the per-request logLevel opt-in from the request envelope and drops the message when the client didn't opt in or the level is below the requested severity; on legacy it honors the session level from logging/setLevel. Because it emits through the request's notify, the modern response upgrades to SSE and the log rides the originating request's stream.
test-servers/configs/subscriptions-legacy-http.json and test-servers/configs/subscriptions-modern-http.json are the resource-subscription era-fork showcase (#1630). Both serve three numbered_resources with subscriptions: true; the legacy one also serves an update_resource tool, and the modern one sets transport.modern: true. Connect to the legacy server, open a resource in the Resources tab and click Subscribe — the client sends resources/subscribe (Network/Protocol view) and the Subscriptions section lists the URI with no stream chrome; call update_resource with that URI and the server updates the content and emits notifications/resources/updated, stamping the subscribed tile's last-updated time. Connect to the modern server with Protocol Era = Modern and the same Subscribe instead sends subscriptions/listen (its filter carries resourceSubscriptions + the resourcesListChanged opt-in) and resolves on notifications/subscriptions/acknowledged; the Subscriptions section then shows the stream-status badge (Connecting… → Listening) in its header, and if the long-lived stream drops it reconnects by re-listing. The modern config deliberately omits update_resource: the SDK's modern leg is stateless/per-request (createMcpHandler(() => createMcpServer(config))), so the tool would run against a throwaway server instance — the content change wouldn't persist for the next resources/read, and its resources/updated wouldn't reach the (separate) listen stream — which is more confusing than useful. The live update-notification round-trip is therefore demonstrated on the legacy (stateful-session) server; the modern server is for the subscribe/listen/badge behavior. (The Inspector's receive path is era-transparent, so a real stateful modern server that routes resources/updated onto the listen stream drives the subscribed tile the same way.)
test-servers/configs/tasks-legacy-http.json and test-servers/configs/tasks-modern-http.json are the Tasks era-fork showcase (#1631). The legacy server advertises capabilities.tasks (tasks: { list, cancel }) with the simple_task / progress_task / elicitation_task presets — connect to it, run one of those tools with Run as task on, and the Tasks tab lists it (populated via tasks/list), polls tasks/get, fetches the payload with the blocking tasks/result, and cancels with tasks/cancel. The modern server sets transport.modern: true and tasksExtension: true, advertising the io.modelcontextprotocol/tasks extension (SEP-2663) and serving the modern_task / modern_input_task tools. Connect with Protocol Era = Modern: the Tasks tab is now gated on the negotiated extension (not capabilities.tasks). Run modern_task as a task — the tools/call returns a CreateTaskResult (resultType: "task", visible in the Protocol/Network tabs), the client polls tasks/get (no tasks/list), and the completed task inlines its result (no blocking tasks/result). Run modern_input_task and the task moves to input_required, surfacing an embedded elicitation through the pending-request modal; answering it sends tasks/update with the inputResponses, and the next poll completes. SDK v2 removed all tasks support and era-gates the tasks/* spec methods out of the modern era on both sides — so the Inspector drives the extension itself (the resultType: "task" frame is rewritten at the transport into a CallToolResult carrying the handle; tasks/get/update/cancel ride a raw-wire request channel with the full modern envelope), and the test server serves tasks/* from an Express interceptor ahead of the SDK handler (the SDK's modern leg would answer them -32601). The Tasks tab's Refresh re-polls the handles already known to the client (modern has no server-side task list).
Building
npm run build # builds all clients: web → cli → tui → launcher
Individual clients: build:web, build:cli, build:tui, build:launcher. The web build produces both the browser SPA (clients/web/dist, Vite) and the Node prod-server runner (clients/web/build, tsup).
Testing & the quality gate
Each client self-validates from its own folder; the root scripts chain them. There is no aggregate root test script — use validate (fast) or coverage (the gate).
| Script | What it does |
|---|---|
npm run validate |
Runs validate:core (the shared core/ format:check + lint gate) first, then per client: format:check + lint + typecheck (cli/tui only) + build + fast unit tests. The quick inner-loop check. |
npm run coverage |
The per-file ≥90% gate (lines/statements/functions/branches) under v8 instrumentation, per client. CI-enforced. For web this also runs the integration project and covers the shared core/ runtime (including core/json and core/client). |
npm run smoke |
End-to-end smokes through the built launcher (--help dispatch + prod cli/tui/web), plus a headless-Chromium boot smoke that runs the prod web bundle and asserts a clean first render (no uncaught error — sync exception or unhandled rejection, how a Node built-in reaching the browser bundle manifests). |
npm run verify:build-gate |
Runs a real vite build with a Node built-in forced into the browser graph and asserts the build fails via the #1769 gate (which turns Vite's browser-externalization warning into a hard error). Guards against the warning phrasing drifting in a Vite bump and silently disabling the gate. Part of npm run ci. |
npm run ci |
Mandatory pre-push command. validate → coverage → verify:build-gate → smoke → Storybook. A true superset of GitHub CI. |
npm run pack:verify |
Publish smoke — see Publishing. |
Per-client scripts exist too (validate:web, coverage:cli, smoke:tui, …), plus root validate:core / format:core for the shared core/ package. Run npm run format before committing — the root format fixes core/ and every client; validate runs the non-fixing format:check and fails CI on any unformatted file.
For the full testing rules — the ≥90% per-file gate, where test files live, the unit vs. integration vs. storybook projects, and the v8 ignore policy — see AGENTS.md.
Publishing
The root @modelcontextprotocol/inspector package ships as one tarball with a single version number — no separate -web / -cli / -tui / -core packages. npm run build builds every client, then prepack runs before npm publish. Runtime dependencies are declared on the root package.json; client builds bundle @inspector/core and externalize npm packages resolved from the root install.
What ships, and the packaging invariants
The root package.json "files" allowlist is the source of truth for the tarball. A few non-obvious entries exist because they are read at runtime or were silently dropped by npm's packlist — do not remove them without re-running npm run pack:verify:
- No source maps. The client bundlers set
sourcemap: false(clients/{cli,tui}/tsup.config.ts,clients/web/tsup.runner.config.ts); Vite and the launcher'stscalready emit none. Maps are ~half the unpacked size and aren't needed at runtime — debug vianpm run devon the source. clients/web/buildships viaclients/web/.npmignore.clients/web/.gitignorelistsbuild/, and npm's packlist honors that nested.gitignoreover the root"files"allowlist — so the prod web-server runner was silently missing from the tarball whileclients/web/distslipped through (its.gitignoreonly listsdist-ssr).clients/web/.npmignoreoverrides the.gitignorefor publishing so bothbuild/(runner) anddist/(SPA) ship. The other clients don't need this — none ship a nested.gitignore.- A single version number, read from the root
package.json. The Inspector ships as one package with one version, so only the rootpackage.jsoncarries aversion— the fourclients/*/package.jsons deliberately have none. Every Node client (CLI, TUI, and the web backend) resolves the version through the sharedreadInspectorVersion()reader incore/node/version.ts, which walks up to the root manifest (always present in the tarball). No clientpackage.jsonis read at runtime, so none needs to ship. The web browser can't read the filesystem; it gets its version from the backend viaGET /api/config(see #1639).
npm run pack:verify — publish smoke against the real tarball
The smoke:* scripts run against the in-repo build tree, which is not the published package. npm run pack:verify (scripts/pack-and-verify.mjs) closes that gap: it builds, npm packs the publishable tarball (asserting no source maps ship and that the runtime-required files are present), installs the tarball into a clean throwaway consumer — a fresh temp directory where it runs a real npm install <tgz> (pulls runtime deps, runs postinstall), exactly as npx @modelcontextprotocol/inspector would — and drives the installed mcp-inspector bin end to end: --help dispatch, a real --cli tools/list over stdio, and a prod --web boot that must serve / from the shipped dist. It catches "works in --dev, breaks under npx …" path/packaging failures. It requires network access (the install pulls deps), so it is a local / release check, not part of the fast validate/ci loop.
Cutting a release
Publishing is automated by two release-gated jobs in .github/workflows/main.yml (github.event_name == 'release', both needs: build):
publish— the npm package. Runsnpm run pack:verifyas the pre-publish gate, asserts the release tag matches the rootpackage.jsonversion, thennpm publish --access public --provenance— a singlenpm publish(v2 is not an npm workspace, so there is no v1-stylepublish-all/--workspaces), with a signed provenance attestation via GitHub OIDC (id-token: write,environment: release,NPM_TOKEN).publish-github-container-registry— the container image (see Docker).
Because there is one version number (only the root package.json has one — the clients carry none, so there is nothing to keep in sync and no check-version step), the release flow is just:
npm version <major|minor|patch> # bumps the root package.json + tags
git push --follow-tags
# then draft & publish a GitHub Release for that tag → triggers `publish`
The release's target commit selects which workflow runs, so this only publishes when a release is cut from a commit carrying this (v2) workflow.
Docker
A container image is published to GHCR (ghcr.io/modelcontextprotocol/inspector, linux/amd64 + linux/arm64) by the release workflow. The Dockerfile is a two-stage build: the first stage installs and npm packs the publishable tarball; the second stage npm install -gs that tarball, so the image ships the exact same artifact as npm, with a clean mcp-inspector bin.
# run the web UI (reads the auth token from the container logs)
docker run --rm -p 6274:6274 ghcr.io/modelcontextprotocol/inspector
# or build the image locally
docker build -t mcp-inspector .
docker run --rm -p 6274:6274 mcp-inspector
The image defaults to --web bound to 0.0.0.0:6274 with browser auto-open disabled; override the args to run another mode (docker run --rm ghcr.io/modelcontextprotocol/inspector --cli …). Pass -e MCP_INSPECTOR_API_TOKEN=… to set a known token (otherwise one is generated and printed in the logs), or -e DANGEROUSLY_OMIT_AUTH=true to disable auth. The image runs as the non-root node user and has a HEALTHCHECK that probes the web UI — it assumes the default --web mode, so add --no-healthcheck when running --cli/--tui (which have no web server).
Contributing — AGENTS.md and CLAUDE.md
AGENTS.md is the contract for changing this codebase, and it applies to humans and AI agents alike. It is not agent-only boilerplate — it holds the project's real conventions: the issue-and-board workflow, branch/label rules, the TypeScript and Mantine/React standards, the testing and coverage requirements, and the mandatory pre-push gate. Read it before making changes, and keep it up to date when you change structure, tooling, or rules.
CLAUDE.md is the entry point the Claude Code agent loads automatically; it simply includes AGENTS.md and this README, so both agents and humans work from the same source of truth. If you use a different agent that reads AGENTS.md, you get the same rules.
A key rule worth surfacing here: all work is issue-driven. Before starting, find or create a tracking issue on the v2 project board; open PRs against v2/main with Closes #<issue>. The exact recipes (labels, board IDs, statuses) are in AGENTS.md.
License
MIT.