maple-onboard

name	maple-onboard
description	Onboard a project to Maple by installing OpenTelemetry traces, logs, and metrics across every app and service in the repo. Triggers on requests like 'install Maple', 'set up Maple', 'add Maple telemetry', 'onboard this repo to Maple', 'instrument with OpenTelemetry for Maple'.

Maple onboarding

Wire OpenTelemetry traces, logs, and metrics into the user's project so telemetry streams to Maple. Cover every app and service in the repo — not just the one the user is currently sitting in.

Prefer native OpenTelemetry APIs and the framework's documented bootstrap over custom helper layers. If a specific stack stumps you, search the OTel docs for that language; don't guess.

Before editing, read the applicable companion skills:

• maple-onboarding-style for general OTel taste, VCS attributes, and smoke checks.
• maple-nextjs-style for Next.js / Vercel apps.
• maple-nodejs-style for plain Node servers (Express, Fastify, Hono, Bun).
• maple-python-style for Python services (FastAPI, Django, Flask).
• maple-effect-style for Effect-based services — Maple's first-class SDK.
• maple-go-style, maple-rust-style, maple-java-style, maple-csharp-style, maple-kotlin-style for those stacks.

If none match (Ruby, Elixir, PHP, plain Deno, …), use maple-onboarding-style as the fallback and consult the upstream OTel docs for that language.

Step 0 — Endpoint and key handling

The OTLP endpoint is always https://ingest.maple.dev and goes inline in the bootstrap code — it's not a secret, no env-var indirection needed.

The ingest API key is project-scoped + write-only — it can only ingest events for one project, can't read anything, can't change settings. Treat it like a Sentry DSN, a PostHog public key, or a Datadog RUM client token: inline it directly in the OTel bootstrap source alongside the endpoint. No .env files, no deploy-target wiring, no process.env.OTEL_EXPORTER_OTLP_HEADERS. The user deploys their code and events flow.

Two paths, no questions asked:

Key in the prompt

If the invoking prompt already contains a Maple ingest key, inline it in every bootstrap file you write. Done — move on to Step 1.

No key

Inline the literal sentinel MAPLE_TEST in the bootstrap source as a placeholder. Maple's ingest gateway accepts it from anyone, returns 200 without forwarding anywhere, so the user's app can boot and exercise the OTel bootstrap path while they go create a real key.

Tell the user briefly at the top of your work: "I'm using MAPLE_TEST as a placeholder so the bootstrap can run. Create a real key at https://app.maple.dev/settings (Settings → API Keys) and search-replace the literal MAPLE_TEST in the files I write to swap it in."

Then keep going. Don't block install on signup.

Step 1 — Map every app/service in the repo

Before instrumenting anything, enumerate what's here. Check workspace manifests (pnpm-workspace.yaml, root package.json workspaces, bun workspaces, go.work, Cargo workspace, Python pyproject.toml workspace setups, apps/* and services/* conventions). Identify each service: web frontend, API, workers, background jobs, CLIs, sample/demo apps, mobile apps, Supabase and/or server functions. Mobile and serverless/edge functions are in scope; do not skip them merely because they are client-side or short-lived. Skip pure type/config packages with no runtime entry point. Do not skip any runnable services or leave them "out of scope": instrument absolutely everything in this run; there may be no follow-up.

Show the user the list before you start, so they can correct it.

Step 2 — For each service, install native OTel and bootstrap

Use the language's native OpenTelemetry SDK. Don't reach for vendor wrappers or hand-rolled helpers when an official package exists. Examples of what "native" means here: @opentelemetry/sdk-node for Node servers, @vercel/otel for normal Next.js/Vercel apps (sdk-node breaks Next's webpack and misses the framework bootstrap), @opentelemetry/sdk-trace-web + browser/mobile-compatible exporters for Vite/SPA/Expo, opentelemetry-instrumentation-* + opentelemetry-sdk for Python, go.opentelemetry.io/otel for Go.

No broad wrapper APIs. Avoid reusable helpers like sendMapleSpan, recordCounter, recordLog, startTelemetrySpan, or withTelemetry. Acquire native tracers/meters/loggers at module scope and use the SDK's own APIs directly. In TypeScript/JavaScript, use the published @maple/otel-helpers withSpan helper for bounded business spans and add @maple/otel-helpers to package.json; this is required when the package can be installed because it avoids expanding a whole function into startActiveSpan plus try / catch / finally. Do not use helpers around provider SDK calls that OpenInference/provider instrumentation can observe directly.

Wire all three signals — traces, logs, metrics. Logs go through OTLP, not just stdout — set up the OTel log bridge for the language so app logs (with their existing log levels and structured fields) carry the active trace_id / span_id automatically. The user's existing logger keeps working; you're just adding an OTLP handler/processor underneath.

Bootstrap rules:

• The bootstrap file must run before any framework imports. Use the language/framework's documented hook (--import flag, instrumentation.ts, top-of-main.py import, etc.).
• Inline the endpoint (https://ingest.maple.dev) and the project's ingest key directly in the bootstrap source. Don't read from process.env.OTEL_EXPORTER_OTLP_* or write any .env files — the key is write-only, and inline configuration removes a whole class of "OTel didn't start because env vars weren't set" deploy failures. (See the framework-specific style skills for the exact shape per stack.)
• Use HTTP OTLP exporters, not gRPC. gRPC pulls in native bindings that break bundlers and complicate containers.
• Use the project's existing package manager (detect via lockfile).
• Prefer idempotent edits. If a config file already exists, edit don't overwrite.
• Set resource attributes on the OTel resource for every service: service.name, service.version, deployment.environment.name, and vcs.repository.url.full — the canonical https URL of the repo (e.g. https://github.com/acme/api). The repo URL is the important one and is fine to hardcode alongside service.name in the SDK init; if the build platform exposes the slug (Vercel VERCEL_GIT_REPO_OWNER/VERCEL_GIT_REPO_SLUG, Railway RAILWAY_GIT_REPO_OWNER/RAILWAY_GIT_REPO_NAME), prefer reading from env. Also set vcs.ref.head.revision (commit SHA) on a best-effort basis from whatever env the runtime already injects (VERCEL_GIT_COMMIT_SHA, RAILWAY_GIT_COMMIT_SHA, GITHUB_SHA, SOURCE_COMMIT, GIT_COMMIT, HEROKU_SLUG_COMMIT, …). Do not shell out to git from the running process. Skipping the SHA is fine, skipping the URL is not. Use the OTel semantic-convention keys exactly — do not invent git.repo / app.repo_url / deployment.commit_sha.

Framework rules:

• Next.js/Vercel: use instrumentation.ts with @vercel/otel registerOTel(...) as the bootstrap. Do not substitute a raw @opentelemetry/sdk-node / NodeSDK bootstrap unless the repo already uses that architecture and you are extending it. Use @opentelemetry/api tracers/meters inside route handlers only where auto-instrumentation is blind.
• Expo/React Native: preserve existing Expo Go / unsupported-runtime guards. In supported builds, call initObservability() before Sentry and before app registration/user code. Inline the endpoint + ingest key in the observability module — no EXPO_PUBLIC_* env vars. The bootstrap reads them straight from constants.
• Supabase Edge Functions / Cloudflare Workers: native Deno / Workers OpenTelemetry can be quirky. Keep the exporter shim tiny, provider-neutral, and OTel-shaped: tracer.startActiveSpan, span.setAttributes, SpanStatusCode, meter.createCounter, histogram.record. For Effect-on-Workers, prefer @maple/effect-cloudflare (see maple-effect-style).
• Python/FastAPI: use native instrumentation such as FastAPIInstrumentor.instrument_app(app) rather than replacing request handling with manual middleware.

Coexist with existing observability vendors. Don't remove Sentry, Datadog, New Relic, Honeycomb, Logtail, Pino transports, etc. OTel sits alongside them. The user explicitly wants both signals flowing during migration; ripping out the incumbent is not your call.

Step 3 — Add custom spans, metrics, and logs around business operations

Auto-instrumentation gets you HTTP in/out, DB queries, framework lifecycle. That's the floor, not the ceiling. Read the project to find the operations a human operator would actually want to see when something looks wrong.

Traces

Wrap every critical business operation with an active span. Auto-instrumented spans are fine where they exist — but if an operation isn't already getting a span, add one.

• Naming: domain.verb (order.process, payment.charge, email.send, agent.run, interview.create, job.<type>).
• Attributes: entity IDs (order.id, user.id, workspace.id, tenant.id), counts, key boolean branch outcomes, model name / provider for LLM calls.
• Record exceptions: span.recordException(err) + span.setStatus({ code: ERROR }) on failure paths. In TS/JS use withSpan from @maple/otel-helpers so the body stays flat instead of nested under try / catch / finally.
• For Python functions with clear boundaries, prefer @tracer.start_as_current_span("operation.name"). Use a context manager when a decorator does not fit. Do not use detached start_span() + manual end() for bounded work.
• Skip trivial getters, pure transforms, internal helpers — anything with no real latency or failure mode.
• Never put PII in attributes (emails, passwords, tokens, full request bodies).

Logs

Make sure logs are structured and carry operation context. Concretely: every log line emitted inside a span should arrive at Maple with trace_id / span_id populated and any structured fields (orderId, userId, etc.) preserved as attributes. Trace/span context may be added natively by the log bridge or integration, or may require additional work.

Use logs for narrative ("starting batch reconcile", "retrying after 3xx") and exceptional events. An error log must only be emitted if the operation cannot recover and manual intervention is required.

Metrics

Cover business + performance + cost. Three categories to look for:

• Business logic counters. Every meaningful state transition: created, started, completed, failed, retried. Per-tenant, per-channel, per-status — low-cardinality dimensions only (never user/order IDs).
• Performance histograms. Latency of operations the user cares about, queue depth, batch sizes, payload sizes. Reuse existing timing instrumentation if the project already has any (time.perf_counter blocks, custom LatencyTrackers, "[TIMING]" log lines) — emit a histogram from those measurements rather than measuring twice.
• Costs — especially LLM costs. If the project calls OpenAI / Anthropic / Google / any LLM provider, prefer provider instrumentation such as OpenInference where available so native SDK calls stay readable. Do not add pricing constants or LLM cost math in product handlers; Maple computes estimated cost centrally in the UI/query layer from captured provider/model/token attributes. Avoid duplicating token counters already captured by provider instrumentation.

Get the meter once at module level, create instruments at module level, increment in the hot path. Don't create a fresh meter per call.

Step 4 — Verify the app still works

Per service:

1. Run the project's own dev or build command (whatever its package.json / pyproject / Makefile already wires up). Confirm it starts cleanly with no errors that trace back to your OTel install. Also run a telemetry bootstrap smoke that imports or starts the app, so provider setup, exporter construction, log bridging, and framework instrumentation all initialize. For a Python server this can be an import/startup command such as uv run python -c 'from app.main import app; print(app.title)'; for Node/Next use the repo's build/start path. For a server, hit at least one route with curl so traffic flows through the instrumentation; choose a route that exercises an instrumented operation when practical, not only a static health route. For a CLI, invoke a real command. Don't ship if the app's own startup is now broken — that's a regression.
2. Confirm telemetry leaves the process. With the inline MAPLE_TEST (or real key) in the bootstrap, OTLP POSTs from the dev server should return 2xx — that proves the bootstrap is reaching the network. As an extra sanity check, you can curl -X POST https://ingest.maple.dev/v1/traces -H "authorization: Bearer MAPLE_TEST" -H "content-type: application/json" -d '{}' and confirm 200. The real signal is the running app's own POSTs succeeding by the time the dev server shuts down — if they don't, the bootstrap is wrong (most often: SDK loaded too late, or shutdown not flushing).

A bootstrap that loads but never POSTs is not a partial success. Fix it before moving on.

Step 5 — Hand-off (final message to the user)

What changed

3–7 short factual bullets covering: packages installed, files created/modified, business spans/metrics added. Per service if changes differed, grouped if uniform. Mention any existing observability vendor (Sentry, Datadog, Logtail, Pino transports, etc.) you intentionally left in place so the coexistence is explicit.

Swap the placeholder if you used one

If MAPLE_TEST is still inline, tell the user:

"The bootstrap currently uses MAPLE_TEST as a placeholder so the install could complete end-to-end. Create a real key at https://app.maple.dev (Settings → API Keys), then search-replace MAPLE_TEST in the files I wrote with your new key and redeploy."

Deploy

Tell the user to deploy as they normally would — push to their hosting platform, run their existing CI, or run locally. There are no env vars to wire and nothing platform-specific to configure: the endpoint and key are inline in the bootstrap, so events start flowing the moment the instrumented code runs.

If the user asks "where do I put the key in production?" — the answer is "you already did, it's in the source you just deployed."

Step 6 — Drive Maple MCP install

Suggest installing the Maple MCP server so the agent (Claude Code, Codex, Cursor, etc.) can query telemetry directly next time the user is debugging — search logs, pull traces, check error rates from inside the chat without context-switching to the dashboard.

For Claude Code (most common — that's where this skill is running), offer to run it for them:

claude mcp add --transport http maple https://api.maple.dev/mcp

This edits the user's Claude Code config. Confirm before running (the user may have a custom MCP scope or want to install elsewhere). If they decline, print the command so they can run it themselves later.

For other agents the user might also use, mention but do not run:

• Codex: codex mcp add maple --url https://api.maple.dev/mcp
• Cursor / others: copy the mcpServers snippet from https://app.maple.dev/mcp.

When done (or skipped), close out with a single line directing the user to deploy their app — they're ready to ship.

Hard rules

• Never modify files outside the project root.
• Never commit, push, or open PRs.
• Inline the ingest key in source. It's a project-scoped, write-only token (think Sentry DSN); env-var indirection just adds deploy-time failure modes for no gain.
• Never remove an existing observability vendor unless the user asks for it.
• Use the project's existing package manager and existing logger.
• Prefer native OTel packages for the language; don't reinvent telemetry plumbing the SDK already provides.
• If the dev/build command errors out because of your instrumentation, that's a failure — fix it or report it, don't paper over it.