v10.38.0

Compare Source

Important Changes

• feat(tanstackstart-react): Auto-instrument request middleware (#​18989)

  The sentryTanstackStart Vite plugin now automatically instruments middleware arrays in createFileRoute(). This captures performance data without requiring manual wrapping with wrapMiddlewaresWithSentry().

Other Changes

• feat: Use v4.8.0 bundler plugins (#​18993)
• feat(browser): Add logs.metrics bundle (#​19020)
• feat(browser): Add replay.logs.metrics bundle (#​19021)
• feat(browser): Add tracing.replay.logs.metrics bundle (#​19039)
• feat(deps): bump import-in-the-middle from 2.0.1 to 2.0.6 (#​19042)
• feat(node): Add AI manual instrumentation exports to Node (#​19063)
• feat(wasm): initialised sentryWasmImages for webworkers (#​18812)
• fix(core): Classify custom AggregateErrors as exception groups (#​19053)
• fix(nextjs): Turn off debugID injection if sourcemaps are explicitly disabled (#​19010)
• fix(react): Avoid String(key) to fix Symbol conversion error (#​18982)
• fix(react): Prevent lazy route handlers from updating wrong navigation span (#​18898)

Work in this release was contributed by @​harshit078. Thank you for your contribution!

v10.37.0

Compare Source

Important Changes

• feat(core): Introduces a new Sentry.setConversationId() API to track multi turn AI conversations across API calls. (#​18909)

  You can now set a conversation ID that will be automatically applied to spans within that scope. This allows you to link traces from the same conversation together.

  import * as Sentry from '@​sentry/node';
  
  // Set conversation ID for all subsequent spans
  Sentry.setConversationId('conv_abc123');
  
  // All AI spans will now include the gen_ai.conversation.id attribute
  await openai.chat.completions.create({...});

  This is particularly useful for tracking multiple AI API calls that are part of the same conversation, allowing you to analyze entire conversation flows in Sentry.
  The conversation ID is stored on the isolation scope and automatically applied to spans via the new conversationIdIntegration.

• feat(tanstackstart-react): Auto-instrument global middleware in sentryTanstackStart Vite plugin (#​18844)

  The sentryTanstackStart Vite plugin now automatically instruments requestMiddleware and functionMiddleware arrays in createStart(). This captures performance data without requiring manual wrapping.

  Auto-instrumentation is enabled by default. To disable it:

  // vite.config.ts
  sentryTanstackStart({
    authToken: process.env.SENTRY_AUTH_TOKEN,
    org: 'your-org',
    project: 'your-project',
    autoInstrumentMiddleware: false,
  });

Other Changes

• feat(core): simplify truncation logic to only keep the newest message (#​18906)
• feat(core): Support new client discard reason invalid (#​18901)
• feat(deps): Bump OpenTelemetry instrumentations (#​18934)
• feat(nextjs): Update default ignore list for sourcemaps (#​18938)
• feat(node): pass prisma instrumentation options through (#​18900)
• feat(nuxt): Don't run source maps related code on Nuxt "prepare" (#​18936)
• feat(replay): Update client report discard reason for invalid sessions (#​18796)
• feat(winston): Add customLevelMap for winston transport (#​18922)
• feat(react-router): Add support for React Router instrumentation API (#​18580)
• fix(astro): Do not show warnings for valid options (#​18947)
• fix(core): Report well known values in gen_ai.operation.name attribute (#​18925)
• fix(node-core): ignore vercel AbortError by default on unhandled rejection (#​18973)
• fix(nuxt): include sentry.config.server.ts in nuxt app types (#​18971)
• fix(profiling): Add platform to envelope item header (#​18954)
• fix(react): Defer React Router span finalization until lazy routes load (#​18881)
• ref(core): rename gen_ai.input.messages.original_length to sentry.sdk_meta.gen_ai.input.messages.original_length (#​18970)
• ref(core): rename gen_ai.request.messages to gen_ai.input.messages (#​18944)
• ref(core): Set system message as separate attribute (#​18978)
• deps: Bump version of sentry-bundler-plugins (#​18972)

Work in this release was contributed by @​sebws, @​harshit078, and @​fedetorre. Thank you for your contributions!

v10.36.0

Compare Source

v10.35.0

Compare Source

Important Changes

• feat(tanstackstart-react): Add sentryTanstackStart vite plugin to manage automatic source map uploads (#​18712)

  You can now configure source maps upload for TanStack Start using the sentryTanstackStart Vite plugin:

  // vite.config.ts
  import { defineConfig } from 'vite';
  import { sentryTanstackStart } from '@​sentry/tanstackstart-react';
  import { tanstackStart } from '@​tanstack/react-start/plugin/vite';
  
  export default defineConfig({
    plugins: [
      sentryTanstackStart({
        authToken: process.env.SENTRY_AUTH_TOKEN,
        org: 'your-org',
        project: 'your-project',
      }),
      tanstackStart(),
    ],
  });

Other Changes

• feat(browser): Add CDN bundle for tracing.replay.feedback.logs.metrics (#​18785)
• feat(browser): Add shim package for logs (#​18831)
• feat(cloudflare): Automatically set the release id when CF_VERSION_METADATA is enabled (#​18855)
• feat(core): Add ignored client report event drop reason (#​18815)
• feat(logs): Add Log exports to browser and node packages (#​18857)
• feat(node-core,bun): Export processSessionIntegration from node-core and add it to bun (#​18852)
• fix(core): Find the correct IP address regardless their case (#​18880)
• fix(core): Check for AI operation id to detect a vercelai span (#​18823)
• fix(ember): Use ES5 syntax in inline vendor scripts (#​18858)
• fix(fetch): Shallow-clone fetch options to prevent mutation (#​18867)

Work in this release was contributed by @​rreckonerr. Thank you for your contribution!

v10.34.0

Compare Source

v10.33.0

Compare Source

Important Changes

• feat(core): Apply scope attributes to metrics (#​18738)

  You can now set attributes on the SDK's scopes which will be applied to all metrics as long as the respective scopes are active. For the time being, only string, number and boolean attribute values are supported.

  Sentry.getGlobalScope().setAttributes({ is_admin: true, auth_provider: 'google' });
  
  Sentry.withScope(scope => {
    scope.setAttribute('step', 'authentication');
  
    // scope attributes `is_admin`, `auth_provider` and `step` are added
    Sentry.metrics.count('clicks', 1, { attributes: { activeSince: 100 } });
    Sentry.metrics.gauge('timeSinceRefresh', 4, { unit: 'hour' });
  });
  
  // scope attributes `is_admin` and `auth_provider` are added
  Sentry.metrics.count('response_time', 283.33, { unit: 'millisecond' });

• feat(tracing): Add Vercel AI SDK v6 support (#​18741)

  The Sentry SDK now supports the Vercel AI SDK v6. Tracing and error monitoring will work automatically with the new version.

• feat(wasm): Add applicationKey option for third-party error filtering (#​18762)

  Adds support for applying an application key to WASM stack frames that can be then used in the thirdPartyErrorFilterIntegration for detection of first-party code.

  Usage:

  Sentry.init({
    integrations: [
      // Integration order matters: wasmIntegration needs to be before thirdPartyErrorFilterIntegration
      wasmIntegration({ applicationKey: 'your-custom-application-key' }), ←───┐
      thirdPartyErrorFilterIntegration({                                      │
        behaviour: 'drop-error-if-exclusively-contains-third-party-frames',   ├─ matching keys
        filterKeys: ['your-custom-application-key'] ←─────────────────────────┘
      }),
    ],
  });

Other Changes

• feat(cloudflare): Support propagateTraceparent (#​18569)
• feat(core): Add ignoreSentryInternalFrames option to thirdPartyErrorFilterIntegration (#​18632)
• feat(core): Add gen_ai.conversation.id attribute to OpenAI and LangGr… (#​18703)
• feat(core): Add recordInputs/recordOutputs options to MCP server wrapper (#​18600)
• feat(core): Support IPv6 hosts in the DSN (#​2996) (#​17708)
• feat(deps): Bump bundler plugins to ^4.6.1 (#​17980)
• feat(nextjs): Emit warning for conflicting treeshaking / debug settings (#​18638)
• feat(nextjs): Print Turbopack note for deprecated webpack options (#​18769)
• feat(node-core): Add isolateTrace option to node-cron instrumentation (#​18416)
• feat(node): Use process.on('SIGTERM') for flushing in Vercel functions (#​17583)
• feat(nuxt): Detect development environment and add dev E2E test (#​18671)
• fix(browser): Forward worker metadata for third-party error filtering (#​18756)
• fix(browser): Reduce number of visibilitystate and pagehide listeners (#​18581)
• fix(browser): Respect tunnel in diagnoseSdkConnectivity (#​18616)
• fix(cloudflare): Consume body of fetch in the Cloudflare transport (#​18545)
• fix(core): Set op on ended Vercel AI spans (#​18601)
• fix(core): Subtract performance.now() from browserPerformanceTimeOrigin fallback (#​18715)
• fix(core): Update client options to allow explicit undefined (#​18024)
• fix(feedback): Fix cases where the outline of inputs were wrong (#​18647)
• fix(next): Ensure inline sourcemaps are generated for wrapped modules in Dev (#​18640)
• fix(next): Wrap all Random APIs with a safe runner (#​18700)
• fix(nextjs): Avoid Edge build warning from OpenTelemetry process.argv0 (#​18759)
• fix(nextjs): Remove polynomial regular expression (#​18725)
• fix(node-core): Ignore worker threads in OnUncaughtException (#​18689)
• fix(node): relax Fastify's setupFastifyErrorHandler argument type (#​18620)
• fix(nuxt): Allow overwriting server-side defaultIntegrations (#​18717)
• fix(pino): Allow custom namespaces for msg and err (#​18597)
• fix(react,solid,vue): Fix parametrization behavior for non-matched routes (#​18735)
• fix(replay): Ensure replays contain canvas rendering when resumed after inactivity (#​18714)
• fix(tracing): add gen_ai.request.messages.original_length attributes (#​18608)
• ref(nextjs): Drop resolve dependency (#​18618)
• ref(react-router): Use snake_case for span op names (#​18617)

Work in this release was contributed by @​G-Rath, @​gianpaj, @​maximepvrt, @​Mohataseem89, @​sebws, and @​xgedev. Thank you for your contributions!

v10.32.1

Compare Source

• fix(cloudflare): Add hono transaction name when error is thrown (#​18529)
• fix(ember): Make implementation field optional (hash routes) (#​18564)
• fix(vercelai): Fix input token count (#​18574)

v10.32.0

Compare Source

Important Changes

• feat(core): Apply scope attributes to logs (#​18184)

  You can now set attributes on the SDK's scopes which will be applied to all logs as long as the respective scopes are active. For the time being, only string, number and boolean attribute values are supported.

  Sentry.geGlobalScope().setAttributes({ is_admin: true, auth_provider: 'google' });
  
  Sentry.withScope(scope => {
    scope.setAttribute('step', 'authentication');
  
    // scope attributes `is_admin`, `auth_provider` and `step` are added
    Sentry.logger.info(`user ${user.id} logged in`, { activeSince: 100 });
    Sentry.logger.info(`updated ${user.id} last activity`);
  });
  
  // scope attributes `is_admin` and `auth_provider` are added
  Sentry.logger.warn('stale website version, reloading page');

• feat(replay): Add Request body with attachRawBodyFromRequest option (#​18501)

  To attach the raw request body (from Request objects passed as the first fetch argument) to replay events, you can now use the attachRawBodyFromRequest option in the Replay integration:

  Sentry.init({
    integrations: [
      Sentry.replayIntegration({
        attachRawBodyFromRequest: true,
      }),
    ],
  });

• feat(tanstackstart-react): Trace server functions (#​18500)

  To enable tracing for server-side requests, you can now explicitly define a server entry point in your application and wrap your request handler with wrapFetchWithSentry.

  // src/server.ts
  import { wrapFetchWithSentry } from '@​sentry/tanstackstart-react';
  import handler, { createServerEntry } from '@​tanstack/react-start/server-entry';
  
  export default createServerEntry(
    wrapFetchWithSentry({
      fetch(request: Request) {
        return handler.fetch(request);
      },
    }),
  );

• feat(vue): Add TanStack Router integration (#​18547)

  The @sentry/vue package now includes support for TanStack Router. Use tanstackRouterBrowserTracingIntegration to automatically instrument pageload and navigation transactions with parameterized routes:

  import { createApp } from 'vue';
  import { createRouter } from '@​tanstack/vue-router';
  import * as Sentry from '@​sentry/vue';
  import { tanstackRouterBrowserTracingIntegration } from '@​sentry/vue/tanstackrouter';
  
  const router = createRouter({
    // your router config
  });
  
  Sentry.init({
    app,
    dsn: '__PUBLIC_DSN__',
    integrations: [tanstackRouterBrowserTracingIntegration(router)],
    tracesSampleRate: 1.0,
  });

Other Changes

• feat(core): Capture initialize attributes on MCP servers (#​18531)
• feat(nextjs): Extract tracing logic from server component wrapper templates (#​18408)
• feat(nextjs): added webpack treeshaking flags as config (#​18359)
• fix(solid/tanstackrouter): Ensure web vitals are sent on pageload (#​18542)

v10.31.0

Compare Source

Important Changes

• feat(browser): Add support for GraphQL persisted operations (#​18505)

The graphqlClientIntegration now supports GraphQL persisted operations (queries). When a persisted query is detected, the integration will capture the operation hash and version as span attributes:

• graphql.persisted_query.hash.sha256 - The SHA-256 hash of the persisted query
• graphql.persisted_query.version - The version of the persisted query protocol

Additionally, the graphql.document attribute format has changed to align with OpenTelemetry semantic conventions. It now contains only the GraphQL query string instead of the full JSON request payload.

Before:

"graphql.document": "{\"query\":\"query Test { user { id } }\"}"

After:

"graphql.document": "query Test { user { id } }"

Other Changes

• feat(node): Support propagateTraceparent option (#​18476)
• feat(bun): Expose spotlight option in TypeScript (#​18436)
• feat(core): Add additional exports for captureException and captureMessage parameter types (#​18521)
• feat(core): Export captureException and captureMessage parameter types (#​18509)
• feat(core): Parse individual cookies from cookie header (#​18325)
• feat(node): Add instrument OpenAI export to node (#​18461)
• feat(nuxt): Bump @sentry/vite-plugin and @sentry/rollup-plugin to 4.6.1 (#​18349)
• feat(profiling): Add support for Node v24 in the prune script (#​18447)
• feat(tracing): strip inline media from messages (#​18413)
• feat(node): Add ESM support for postgres.js instrumentation (#​17961)
• fix(browser): Stringify span context in linked traces log statement (#​18376)
• fix(google-cloud-serverless): Move @​types/express to optional peerDeps (#​18452)
• fix(node-core): passthrough node-cron context (#​17835)
• fix(tanstack-router): Check for fromLocation existence before reporting pageload (#​18463)
• fix(tracing): add system prompt, model to google genai (#​18424)
• fix(tracing): Set span operations for AI spans with model ID only (#​18471)
• ref(browser): Improve profiling debug statement (#​18507)

Work in this release was contributed by @​sebws and @​TBeeren. Thank you for your contributions!

Bundle size 📦

v10.30.0

Compare Source

• feat(nextjs): Deprecate Webpack top-level options (#​18343)
• feat(node): Capture scope when event loop blocked (#​18040)
• fix(aws-serverless): Remove hyphens from AWS-lambda origins (#​18353)
• fix(core): Parse method from Request object in fetch (#​18453)
• fix(react): Add transaction name guards for rapid lazy-route navigations (#​18346)

Bundle size 📦

v10.29.0

Compare Source

Important Changes

• feat(solid|solidstart): Bump accepted @​solidjs/router range (#​18395)

We expanded the supported version range for @solidjs/router to include 0.14.x and 0.15.x versions.

Other Changes

• fix(logs): Add support for msg in pino integration (#​18389)
• fix(node): Include system message in anthropic-ai messages span (#​18332)
• fix(tracing): Add missing attributes in vercel-ai spans (#​18333)

Bundle size 📦