LibreChat/api/server/index.js
Danny Avila ee8c0abe2d
🪝 feat: Execute Agent Plugin Command Hooks (#14755)
* 🪝 feat: Execute Agent Plugin Command Hooks

Implement the missing PluginHookExecutor boundary so deployment plugins'
ai.librechat/hooks/hooks.json documents execute instead of loading inert:

- Command executor runs handlers as child processes outside the API
  process: Claude-shaped JSON payload on stdin, exit 0 + JSON stdout as
  sanitized hook output, exit 2 blocks with stderr as the reason, minimal
  allowlisted environment plus PLUGIN_ROOT/PLUGIN_DATA, abort-signal kill
- Plugin loading carries the parsed hooks document on the contribution and
  threads hookCapabilities from startup, gated on the operator opt-in
  DEPLOYMENT_PLUGIN_HOOKS (off by default: parsed-but-inert with warning)
- Runs register every ready plugin hook onto the per-run HookRegistry after
  internal policy hooks, with once-per-conversation SessionStart dedup

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Harden Plugin Hook Execution Boundary

Address CI and Codex/Copilot review findings on #14755:

- Break the agents -> plugins import cycle: the run seam now reads a
  PluginHookSource wired at startup (mirrors the tool-approval registry)
- Tighten plugin ask decisions to deny unless the run has HITL wiring,
  so an un-resumable interrupt can never strand OpenAI-compatible callers
- Scope cross-run dedup keys by authenticated user and handler identity:
  caller-supplied conversation ids cannot collide across principals, and
  sibling SessionStart handlers all fire; once handlers persist across runs
- Replace a literal NUL byte in source with an escape (file diffed binary)
- Kill the whole detached process group on abort, not just the shell
- Map exit 2 on events without a decision channel to preventContinuation
- Reserve PLUGIN_ROOT/PLUGIN_DATA against allowlist overrides, quote
  PowerShell args, cap captured output by bytes with one-pass decoding,
  and serialize payloads inside the executor's error boundary
- Fix import ordering flagged by the static checks

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Close Plugin Hook Policy and Namespace Gaps

Address the second Codex review round on #14755:

- Drop updatedInput from plugin command outputs: hooks in one dispatch all
  receive the original arguments, so a plugin rewrite would reach the tool
  without the approval policy re-evaluating it (host-only now)
- Translate Claude tool aliases (Bash/Write/Edit/Read) to LibreChat runtime
  names in matchers, with reverse payload mapping, so Claude-authored guards
  fire instead of planning ready and never matching
- Key once-only state by declaration position as well as handler contents,
  so sibling declarations with identical handlers stay independent
- Thread sessionStartSource through createRun and mark the HITL resume
  rebuild as 'resume', so SessionStart matchers see the real lifecycle

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Translate Regex-Form Claude Tool Aliases

Address the third Codex review round on #14755: alias translation now
substitutes word-bounded tokens, covering regex matchers like ^Bash$ and
^(Write|Edit)$ that the exact-token pass left registered against Claude
names and silently never firing. A regex whose alias sits inside a
character class or escape is rejected as unmapped so it fails loudly at
plan time instead of never running.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Scope Alias Translation and Reuse Load-Time Plans

Address the fourth Codex review round on #14755:

- Add the WebSearch -> web_search alias so Claude-authored web-search
  guards fire against the LibreChat built-in
- Apply alias translation only to tool-name events; a StopFailure matcher
  like ^Bash failed$ stays untouched and keeps matching the error text
- Reuse each plugin's load-time hook plan at run registration instead of
  re-planning up to 512 handlers on every chat turn

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Translate Aliased Tool Inputs and Harden Hook Domains

- Present aliased tool inputs under Claude field names (file_path,
  old_string, new_string, including nested edits), so Write/Edit/Read
  guards see the fields they check instead of silently allowing
- Derive the alias table from canonical tool-name definitions
  (BashExecutionToolDefinition, CREATE_FILE_TOOL_NAME, Tools.web_search)
  instead of a parallel hand-authored table
- Reject matchers naming Claude built-ins with no runtime equivalent
  (Task, Glob, Grep, WebFetch, ...) as unmapped at plan time instead of
  registering guards that never fire
- Replace per-event Sets and Stop special-cases with an exhaustive
  EVENT_TRAITS record over HookEvent, so new engine events demand
  explicit semantics at compile time
- Move cross-run once-state behind a PluginHookOnceStore seam with a
  least-recently-marked memory default: active conversations refresh
  their keys each turn, so capacity eviction can no longer re-fire a
  conversation that is still in use; the seam admits a shared-cache
  store for multi-replica deployments
- Gate portable-only command handlers at plan time on Windows via a new
  supportsHandler capability (commandWindows or shell powershell
  required) instead of spawning bash that cannot exist
- Kill Windows hook process trees with taskkill /t on abort
- Require declaration indices on execution requests, stamped from the
  plan instead of defaulted at execution time

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Keep Group SIGKILL Escalation Armed After Wrapper Exit

An aborted hook whose descendant ignores SIGTERM could leak that
descendant: the wrapper shell's exit fired close, which cancelled the
scheduled group SIGKILL. The escalation timer is now never cancelled —
it is unref'd and killTree already tolerates a vanished process group,
so a redundant late sweep is harmless while a surviving descendant is
reliably killed at the grace deadline. killGraceMs is configurable on
CommandExecutorOptions, with a regression test driving a trap-protected
descendant past the wrapper's exit.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Scope Once Retention by Conversation and Reject Clear Source

- Restructure the once store around conversation scopes: registration
  touches the scope every run, so rarely-matching once handlers keep
  their keys while the conversation is active; eviction removes whole
  idle conversations (capacity counts conversations, not keys)
- Reject SessionStart matchers naming the clear lifecycle source at
  plan time — no LibreChat run-construction path emits clear, so the
  handler would plan ready and never fire; wildcard warning text now
  reflects the sources that actually occur
- Make the SIGKILL-escalation regression test real: the surviving
  descendant redirects its stdio away from the captured pipes so the
  wrapper's close fires while it is still alive, exercising the
  window a close-time cancellation would leak

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Bound Alias Tokens by Tool-Name Characters and Host Shells

- Translate Claude aliases (and reject unsupported built-ins) only when
  delimited by characters that cannot appear in a runtime tool name:
  action tool names preserve hyphens, so an alias embedded in a longer
  name like deploy-Bash-v2_action_example_com stays the literal tool
  name instead of being rewritten into a matcher that never fires
- Reject PowerShell-only command handlers on POSIX hosts at plan time
  (and skip them at runtime): bash cannot run PowerShell syntax, so the
  guard would fail open; a handler with both variants still runs its
  portable command
- Handle rejected asynchronous once-store calls: a failed touch logs
  instead of raising an unhandled rejection during run construction,
  and a failed markOnce lookup fails open per the store's documented
  over-fire direction

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Probe Group Liveness Before Cancelled or Delivered SIGKILL

The never-cancelled escalation timer could signal a recycled
process-group id when an aborted hook's whole tree exits early in the
grace window. Escalation now probes the group with signal 0: close
cancels the timer only when the group is verifiably empty, and the
deadline re-probes before delivering the group SIGKILL, so surviving
descendants are still reaped while a fully-dead group never receives a
blind late signal. The residual probe-to-signal race is documented as
irreducible without pidfd support.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Gate Windows Escalation on Root-Process Liveness

Windows taskkill /t walks the tree from the root process, so once Node
observes the root's exit an escalation pass can reap nothing and a late
forced taskkill could only hit a recycled PID. The liveness gate is now
platform-aware in one helper: POSIX probes the process group with
signal 0, Windows checks the root's observed exit state, and both the
close-time cancellation and the deadline delivery consult it — no
platform retains a blind late signal. Orphaned SIGTERM-ignoring
descendants on Windows are documented as the platform limitation they
are without Job Objects.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Scope Payload Namespace to Declarations and Reap Stray Workers

- Reverse name/input translation now applies only to declarations whose
  matcher actually required Claude-alias translation: the plan records
  requiresToolNameTranslation per entry, so a native-authored matcher
  like ^create_file$ receives native tool names and fields instead of
  Claude-shaped payloads its guard never expected
- Coordinate the two dedup layers via a shouldExecute gate on the
  executor: a declaration suppressed by spent once-state declines
  before claiming the per-input dedup slot, so an identical handler
  under an overlapping matcher can still claim it and fire its own
  independent once-key instead of being permanently shadowed
- Reap process groups that outlive a successful hook: a backgrounded
  worker left running after normal wrapper exit gets the same
  term-then-escalate sequence an abort uses, since unsupported async
  handlers mean no lifecycle owns such processes

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🧰 chore: Vendor Pocock Codebase-Design and Architecture Skills

Adds mattpocock/skills engineering/codebase-design and
engineering/improve-codebase-architecture (MIT, license included) under
.claude/skills so future sessions share the deep-module vocabulary
(module, interface, depth, seam, adapter, leverage, locality) and the
architecture-review process. Force-added past the /.claude/ gitignore
deliberately; relocate if project skills should live elsewhere.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 refactor: Extract Process-Tree Reaping Into a Reaper Module

Tree lifecycle — five of the last seven review findings — lived as
event-handler wiring inside runCommand with its invariants in comments.
It now sits behind a two-method seam: createReaper(child, graceMs)
exposes reap() and onClose(), hiding the term-grace-escalate state
machine, the per-platform liveness gates, the recycled-id guards, and
the clean-exit sweep. The executor shrinks to capture-and-parse, and
the reaper is unit-tested directly with real process trees through its
own interface instead of only via whole-executor integration runs.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Scope Translation Per Alternative and Sweep at Root Exit

- Track which runtime tool names alias translation produced, so a
  mixed-namespace matcher like Bash|create_file presents Claude-shaped
  payloads only for bash_tool invocations while the natively-authored
  create_file alternative keeps native names and fields; a capability
  omitting the produced-names list keeps declaration-wide translation
- Sweep the process tree at root exit as well as close: a backgrounded
  descendant holding the captured pipes delays close until it dies, so
  the exit-time sweep terminates it promptly instead of stalling the
  hook until its timeout aborts
- Pass the primary agent's resolved model and identity into the plugin
  hook context, so SessionStart payloads carry model and agent_type
  instead of always omitting them

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Default Wildcard Declarations to the Document Namespace

- Matcherless (or wildcard) tool-payload declarations now inherit the
  hook document's Claude namespace: with no alternatives to carry
  namespace evidence, the plan marks them for declaration-wide reverse
  translation, so a wildcard guard inspecting standard Claude names and
  fields sees Write/file_path instead of silently failing open on
  native payloads; PostToolBatch entries translate the same way
- Recognize aliases delimited by regex metacharacters: dots leave the
  tool-name boundary class (runtime names never contain them — action
  ids underscore domain dots), so ^Bash.*$ translates to ^bash_tool.*$
  instead of registering a guard that never fires
- Expand Claude's ${CLAUDE_PLUGIN_ROOT} spelling in hook commands and
  export it in the child environment alongside PLUGIN_ROOT
- Scope SessionStart once-keys by lifecycle source, so a startup firing
  no longer suppresses the conversation's resume rebuild

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Normalize Claude Structured Hook Output

Stock Claude hooks return decisions under hookSpecificOutput
(permissionDecision/permissionDecisionReason), surface context there,
and use continue:false plus the legacy approve/block decisions — none
of which the sanitizer's native field names recognized, so a guard that
works in Claude silently allowed in LibreChat. Parsed JSON now passes
through a dialect normalizer first: hookSpecificOutput fields map to
decision/reason/additionalContext, continue:false becomes
preventContinuation, approve becomes allow, and block becomes deny on
events that block by denying. Native fields win when both dialects
appear, and the ask-to-deny gate applies to the Claude dialect too.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Validate Native Decisions and Slim Once Keys

- Strip malformed native output fields before the dialect merge, so a
  placeholder like {"decision":null} can no longer suppress a valid
  Claude permissionDecision into a silent allow; only recognized
  decision tokens take precedence
- Preserve the caller's working directory in hook payloads: cwd now
  reports the run's session context instead of the plugin installation
  path, which commands already receive as PLUGIN_ROOT and which the
  executor still uses as each process's working directory
- Store a compact sha256 digest instead of the full serialized handler
  in once keys: declarations may carry 32 KB commands and 256 args, and
  the previous key embedded them in every retained conversation scope

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

* 🪝 fix: Validate Decisions Per Event Channel and Control Post-Tool Blocks

- Accept native decision tokens only from the target event's own
  vocabulary: "continue" is valid on Stop but malformed on a tool
  event, where it previously survived validation, blocked the Claude
  dialect merge, and was then dropped by sanitization into a silent
  allow
- Translate a structured "block" on events with no deny channel
  (PostToolUse, PostToolUseFailure, and the other prevent-trait events)
  into preventContinuation with the block reason as stopReason, instead
  of discarding it and returning a reason that controls nothing
- Document why LibreChat runs supply no payload cwd: tool paths address
  a remote code-execution sandbox rather than the API host where hook
  commands run, so no host directory describes the run

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MWXQZD2WzeAsvee4eRdWWc

---------

Co-authored-by: Claude <noreply@anthropic.com>
2026-08-12 13:21:15 -04:00

512 lines
18 KiB
JavaScript

require('../config/credentials');
const telemetry = require('./telemetry');
const fs = require('fs');
const path = require('path');
require('module-alias')({ base: path.resolve(__dirname, '..') });
const cors = require('cors');
const axios = require('axios');
const express = require('express');
const passport = require('passport');
const compression = require('compression');
const cookieParser = require('cookie-parser');
const mongoSanitize = require('express-mongo-sanitize');
const { logger, runAsSystem } = require('@librechat/data-schemas');
const {
isEnabled,
apiNotFound,
createMetrics,
ErrorController,
memoryDiagnostics,
performStartupChecks,
handleJsonParseError,
GenerationJobManager,
QUERY_DEVTOOLS_HEADER,
createStreamServices,
agentStartupIngressMiddleware,
agentStartupTelemetryMiddleware,
initializeFileStorage,
initializeDeploymentSkills,
initializeDeploymentPlugins,
getDeploymentPluginSkills,
getDeploymentPluginHookCapabilities,
registerDeploymentPluginHooks,
hasDeploymentPluginHooks,
setPluginHookSource,
loadToolApprovalHooks,
maybeInjectQueryDevtoolsBootstrap,
preAuthTenantMiddleware,
requestContextMiddleware,
registerShutdownTask,
configureServerTimeouts,
setupGracefulShutdown,
updateInterfacePermissions,
configureMessageFilterRegexValidator,
configureFileConfigRegexEngine,
waitForKeyvRedisClient,
} = require('@librechat/api');
const { connectDb, indexSync } = require('~/db');
const {
updateAccessPermissions,
sweepOrphanedPreviews,
getRoleByName,
seedDatabase,
} = require('~/models');
const initializeOAuthReconnectManager = require('./services/initializeOAuthReconnectManager');
const { capabilityContextMiddleware } = require('./middleware/roles/capabilities');
const createValidateImageRequest = require('./middleware/validateImageRequest');
const { initializeGitHubSkillSync } = require('./services/Skills/sync');
const { jwtLogin, ldapLogin, passportLogin } = require('~/strategies');
const { startExpiredFileSweep } = require('./services/Files/process');
const { checkMigrations } = require('./services/start/migration');
const optionalJwtAuth = require('./middleware/optionalJwtAuth');
const initializeMCPs = require('./services/initializeMCPs');
const configureSocialLogins = require('./socialLogins');
const createSpaFallback = require('./utils/fallback');
const { getAppConfig } = require('./services/Config');
const staticCache = require('./utils/staticCache');
const noIndex = require('./middleware/noIndex');
const routes = require('./routes');
/** Route admin file-config MIME patterns through a linear-time engine (ReDoS-safe) on upload. */
configureFileConfigRegexEngine();
/** Reject messageFilter PII patterns the RE2 runtime engine cannot compile, at config load. */
configureMessageFilterRegexValidator();
const { PORT, HOST, ALLOW_SOCIAL_LOGIN, DISABLE_COMPRESSION, TRUST_PROXY } = process.env ?? {};
// Allow PORT=0 to be used for automatic free port assignment
const port = isNaN(Number(PORT)) ? 3080 : Number(PORT);
const host = HOST || 'localhost';
const trusted_proxy = Number(TRUST_PROXY) || 1; /* trust first proxy by default */
const app = express();
let serverReady = false;
const SERVER_NOT_READY_CODE = 'SERVER_NOT_READY';
const CHAT_START_RETRY_AFTER_SECONDS = '1';
const rejectChatStartsUntilReady = (req, res, next) => {
if (serverReady || req.method !== 'POST' || req.path === '/abort') {
return next();
}
res.set('Retry-After', CHAT_START_RETRY_AFTER_SECONDS);
return res.status(503).json({
code: SERVER_NOT_READY_CODE,
error: 'Server is still starting. Please retry shortly.',
});
};
const configureGenerationStreams = () => {
const streamServices = createStreamServices();
GenerationJobManager.configure({
...streamServices,
cleanupOnComplete: !isEnabled(process.env.STREAM_KEEP_COMPLETED_JOBS),
});
GenerationJobManager.initialize();
// Stop active generations and close their SSE streams while the HTTP server drains.
registerShutdownTask(
'generation job manager prepare',
() => GenerationJobManager.prepareForShutdown(),
{
phase: 'pre-drain',
priority: 100,
},
);
// Tear down stream resources before shared caches and telemetry exporters shut down.
registerShutdownTask('generation job manager', () => GenerationJobManager.destroy(), {
priority: 100,
});
};
const startServer = async () => {
await waitForKeyvRedisClient();
const { metricsMiddleware, metricsRouter } = createMetrics();
if (!process.env.METRICS_SECRET) {
logger.warn('[metrics] METRICS_SECRET is not set - /metrics will return 401 for all requests');
}
if (typeof Bun !== 'undefined') {
axios.defaults.headers.common['Accept-Encoding'] = 'gzip';
}
await connectDb();
logger.info('Connected to MongoDB');
indexSync().catch((err) => {
logger.error('[indexSync] Background sync failed:', err);
});
app.disable('x-powered-by');
app.set('trust proxy', trusted_proxy);
if (isEnabled(process.env.TENANT_ISOLATION_STRICT)) {
logger.warn(
'[Security] TENANT_ISOLATION_STRICT is active. Ensure your reverse proxy strips or sets ' +
'the X-Tenant-Id header — untrusted clients must not be able to set it directly.',
);
}
await runAsSystem(seedDatabase);
/* Recover stuck `status: 'pending'` records from a crash mid-render.
* `runAsSystem` is required — `File` is tenant-isolated and strict
* mode rejects unscoped queries. Lazy sweep in the preview endpoint
* covers anything younger than the boot cutoff. */
runAsSystem(sweepOrphanedPreviews).catch((err) => {
logger.error('[sweepOrphanedPreviews] Background sweep failed:', err);
});
const appConfig = await getAppConfig({ baseOnly: true });
initializeFileStorage(appConfig);
const projectRoot = path.resolve(__dirname, '../..');
// Plugin hooks execute only when the operator opts in via DEPLOYMENT_PLUGIN_HOOKS;
// without it, declared hook documents load as parsed-but-inert with a warning.
await initializeDeploymentPlugins({
projectRoot,
hookCapabilities: getDeploymentPluginHookCapabilities(),
});
// Hand the run seam its plugin-hook source without a packages/api-internal
// agents -> plugins import (see agents/hooks/source.ts).
setPluginHookSource({
hasHooks: hasDeploymentPluginHooks,
register: registerDeploymentPluginHooks,
});
await initializeDeploymentSkills({
projectRoot,
additionalSkills: getDeploymentPluginSkills(),
});
initializeGitHubSkillSync(appConfig);
startExpiredFileSweep({ appConfig, loadAppConfig: getAppConfig });
// Register any programmatic tool-approval policy hooks declared in
// `endpoints.agents.toolApproval.hooks`. Honor the `enabled` kill switch: when tool
// approval is off we pass no hooks, so a disabled endpoint imports/runs nothing (and any
// previously loaded batch is unregistered). Hooks are read from the BASE config only —
// they register once, process-wide; per-user/tenant differences belong inside the hook
// (via its context), not in per-override module lists.
const toolApproval = appConfig?.endpoints?.agents?.toolApproval;
await loadToolApprovalHooks(toolApproval?.enabled ? toolApproval.hooks : undefined, {
basePath: path.resolve(__dirname, '../..'),
});
await runAsSystem(async () => {
await performStartupChecks(appConfig);
await updateInterfacePermissions({ appConfig, getRoleByName, updateAccessPermissions });
});
const indexPath = path.join(appConfig.paths.dist, 'index.html');
let indexHTML = fs.readFileSync(indexPath, 'utf8');
// In order to provide support to serving the application in a sub-directory
// We need to update the base href if the DOMAIN_CLIENT is specified and not the root path
if (process.env.DOMAIN_CLIENT) {
const clientUrl = new URL(process.env.DOMAIN_CLIENT);
const baseHref = clientUrl.pathname.endsWith('/')
? clientUrl.pathname
: `${clientUrl.pathname}/`;
if (baseHref !== '/') {
logger.info(`Setting base href to ${baseHref}`);
indexHTML = indexHTML.replace(/base href="\/"/, `base href="${baseHref}"`);
}
}
const sendIndexHtml = (req, res) => {
res.set({
'Cache-Control': process.env.INDEX_CACHE_CONTROL || 'no-cache, no-store, must-revalidate',
Pragma: process.env.INDEX_PRAGMA || 'no-cache',
Expires: process.env.INDEX_EXPIRES || '0',
});
res.vary(QUERY_DEVTOOLS_HEADER);
const lang = req.cookies.lang || req.headers['accept-language']?.split(',')[0] || 'en-US';
const saneLang = lang.replace(/"/g, '&quot;');
let updatedIndexHtml = indexHTML.replace(/lang="en-US"/g, `lang="${saneLang}"`);
updatedIndexHtml = maybeInjectQueryDevtoolsBootstrap(updatedIndexHtml, req);
res.type('html');
res.send(updatedIndexHtml);
};
app.get('/health', (_req, res) => res.status(200).send('OK'));
app.get('/livez', (_req, res) => res.status(200).send('OK'));
app.get('/readyz', (_req, res) => {
if (!serverReady) {
return res.status(503).send('NOT_READY');
}
return res.status(200).send('OK');
});
/* Middleware */
app.use(requestContextMiddleware);
app.use('/api/agents/chat', agentStartupIngressMiddleware);
app.use(metricsMiddleware);
app.use(noIndex);
app.use(express.json({ limit: '3mb' }));
app.use(express.urlencoded({ extended: true, limit: '3mb' }));
app.use(handleJsonParseError);
/**
* Express 5 Compatibility: Make req.query writable for mongoSanitize
* In Express 5, req.query is read-only by default, but express-mongo-sanitize needs to modify it
*/
app.use((req, _res, next) => {
Object.defineProperty(req, 'query', {
...Object.getOwnPropertyDescriptor(req, 'query'),
value: req.query,
writable: true,
});
next();
});
app.use(mongoSanitize());
app.use(cors());
app.use(cookieParser());
if (!isEnabled(DISABLE_COMPRESSION)) {
app.use(compression());
} else {
console.warn('Response compression has been disabled via DISABLE_COMPRESSION.');
}
app.get('/index.html', sendIndexHtml);
app.use(staticCache(appConfig.paths.dist));
app.use(staticCache(appConfig.paths.fonts));
app.use(staticCache(appConfig.paths.assets));
if (telemetry.enabled) {
app.use(telemetry.telemetryMiddleware);
}
app.use('/api/agents/chat', agentStartupTelemetryMiddleware);
if (!ALLOW_SOCIAL_LOGIN) {
console.warn('Social logins are disabled. Set ALLOW_SOCIAL_LOGIN=true to enable them.');
}
/* OAUTH */
app.use(passport.initialize());
passport.use(jwtLogin());
passport.use(passportLogin());
/* LDAP Auth */
if (process.env.LDAP_URL && process.env.LDAP_USER_SEARCH_BASE) {
passport.use(ldapLogin);
}
if (isEnabled(ALLOW_SOCIAL_LOGIN)) {
await configureSocialLogins(app);
}
/* Per-request capability cache — must be registered before any route that calls hasCapability */
app.use(capabilityContextMiddleware);
/* Pre-auth tenant context for unauthenticated routes that need tenant scoping.
* The reverse proxy / auth gateway sets `X-Tenant-Id` header for multi-tenant deployments. */
app.use('/oauth', preAuthTenantMiddleware, routes.oauth);
/* API Endpoints */
app.use('/api/auth', preAuthTenantMiddleware, routes.auth);
app.use('/api/admin', routes.adminAuth);
app.use('/api/admin/config', routes.adminConfig);
app.use('/api/admin/langfuse', routes.adminLangfuse);
app.use('/api/admin/grants', routes.adminGrants);
app.use('/api/admin/groups', routes.adminGroups);
app.use('/api/admin/roles', routes.adminRoles);
app.use('/api/admin/skills', routes.adminSkills);
app.use('/api/admin/users', routes.adminUsers);
app.use('/api/admin/audit-log', routes.adminAuditLog);
app.use('/api/actions', routes.actions);
app.use('/api/keys', routes.keys);
app.use('/api/api-keys', routes.apiKeys);
app.use('/api/user', routes.user);
app.use('/api/search', routes.search);
app.use('/api/messages', routes.messages);
app.use('/api/convos', routes.convos);
app.use('/api/presets', routes.presets);
app.use('/api/projects', routes.projects);
app.use('/api/prompts', routes.prompts);
app.use('/api/skills', routes.skills);
app.use('/api/categories', routes.categories);
app.use('/api/endpoints', routes.endpoints);
app.use('/api/balance', routes.balance);
app.use('/api/models', routes.models);
app.use('/api/config', preAuthTenantMiddleware, optionalJwtAuth, routes.config);
app.use('/api/assistants', routes.assistants);
app.use('/api/files', await routes.files.initialize());
app.use('/images/', createValidateImageRequest(appConfig.secureImageLinks), routes.staticRoute);
app.use('/api/share', preAuthTenantMiddleware, routes.share);
app.use('/api/roles', routes.roles);
app.use('/api/agents/chat', rejectChatStartsUntilReady);
app.use('/api/agents', routes.agents);
app.use('/api/banner', routes.banner);
app.use('/api/memories', routes.memories);
app.use('/api/permissions', routes.accessPermissions);
app.use('/api/tags', routes.tags);
app.use('/api/mcp', routes.mcp);
app.use('/api/rum', routes.rum);
app.use('/metrics', metricsRouter);
/** 404 for unmatched API routes */
app.use('/api', apiNotFound);
/** SPA fallback - serve index.html for all unmatched routes */
app.use(createSpaFallback(sendIndexHtml));
/** Record trace errors before the final error controller. */
if (telemetry.enabled) {
app.use(telemetry.telemetryErrorMiddleware);
}
/** Error handler (must be last - Express identifies error middleware by its 4-arg signature) */
app.use(ErrorController);
configureGenerationStreams();
const server = app.listen(port, host, async (err) => {
if (err) {
logger.error('Failed to start server:', err);
process.exit(1);
}
if (host === '0.0.0.0') {
logger.info(
`Server listening on all interfaces at port ${port}. Use http://localhost:${port} to access it`,
);
} else {
logger.info(`Server listening at http://${host == '0.0.0.0' ? 'localhost' : host}:${port}`);
}
/**
* The listen callback is async, so any rejection from these awaits would
* otherwise be detached from `startServer().catch(...)` (which only
* catches errors that happen before `app.listen`). Without explicit
* handling, the global `unhandledRejection` handler would swallow init
* failures and leave the server listening but only partially
* initialized — passing liveness checks while serving broken requests.
*/
try {
await runAsSystem(async () => {
await initializeMCPs();
await initializeOAuthReconnectManager();
});
await checkMigrations();
const inspectFlags = process.execArgv.some((arg) => arg.startsWith('--inspect'));
if (inspectFlags || isEnabled(process.env.MEM_DIAG)) {
memoryDiagnostics.start();
}
serverReady = true;
logger.info('Server readiness checks passing.');
} catch (initErr) {
serverReady = false;
logger.error('Post-listen initialization failed:', initErr);
process.exit(1);
}
});
configureServerTimeouts(server);
logger.info('HTTP server timeout configuration', {
keepAliveTimeout: server.keepAliveTimeout,
keepAliveTimeoutBuffer: server.keepAliveTimeoutBuffer,
headersTimeout: server.headersTimeout,
requestTimeout: server.requestTimeout,
});
setupGracefulShutdown(server);
};
/**
* Boot rejections (e.g. `connectDb`, `getAppConfig`, `performStartupChecks`)
* must remain fail-fast: a half-initialized process with no listening HTTP
* server should die immediately so the orchestrator restarts it, instead of
* being kept alive by the `unhandledRejection` handler below until the
* liveness probe eventually times out. Mirrors the pattern in
* `experimental.js`.
*/
startServer().catch((err) => {
logger.error('Failed to start server:', err);
process.exit(1);
});
let messageCount = 0;
process.on('uncaughtException', (err) => {
if (!err.message.includes('fetch failed')) {
logger.error('There was an uncaught error:', err);
}
if (err.message && err.message?.toLowerCase()?.includes('abort')) {
logger.warn('There was an uncatchable abort error.');
return;
}
if (err.message.includes('GoogleGenerativeAI')) {
logger.warn(
'\n\n`GoogleGenerativeAI` errors cannot be caught due to an upstream issue, see: https://github.com/google-gemini/generative-ai-js/issues/303',
);
return;
}
if (err.message.includes('fetch failed')) {
if (messageCount === 0) {
logger.warn('Meilisearch error, search will be disabled');
messageCount++;
}
return;
}
if (err.message.includes('OpenAIError') || err.message.includes('ChatCompletionMessage')) {
logger.error(
'\n\nAn Uncaught `OpenAIError` error may be due to your reverse-proxy setup or stream configuration, or a bug in the `openai` node package.',
);
return;
}
if (err.stack && err.stack.includes('@librechat/agents')) {
logger.error(
'\n\nAn error occurred in the agents system. The error has been logged and the app will continue running.',
{
message: err.message,
stack: err.stack,
},
);
return;
}
if (isEnabled(process.env.CONTINUE_ON_UNCAUGHT_EXCEPTION)) {
logger.error('Unhandled error encountered. The app will continue running.', {
name: err?.name,
message: err?.message,
stack: err?.stack,
});
return;
}
process.exit(1);
});
/**
* Unhandled promise rejection handler.
*
* Node 15+ terminates the process by default when a promise rejection is
* unhandled. MCP OAuth reconnect storms and streamable-HTTP transport resets
* can produce transient fire-and-forget rejections (ECONNRESET, token refresh
* races) that are recoverable — the server should log and keep serving other
* requests rather than silently crash under load.
*
* Non-Error reasons are forwarded as-is so structured payloads (e.g.
* `{ code: "ECONNRESET", errno: -104 }`) survive instead of being collapsed to
* "[object Object]" by `String()`.
*/
process.on('unhandledRejection', (reason) => {
if (reason instanceof Error) {
logger.error('Unhandled promise rejection. The app will continue running.', {
name: reason.name,
message: reason.message,
stack: reason.stack,
cause: reason.cause,
});
return;
}
logger.error('Unhandled promise rejection. The app will continue running.', { reason });
});
/** Export app for easier testing purposes */
module.exports = app;