See how your AI tools are actually used.
Cross-tool AI assistant observability. Skills invoked, tokens consumed, session quality — all pushed to your infrastructure. Crucible never sees the data.
The enterprise visibility gap
No idea which skills are used
You sync 20 skills to 50 developers. Which ones are actually invoked? Which are ignored? No way to know.
No cost tracking
Uber spent their entire 2026 AI budget in 4 months. How do you measure value per team, per repo, per assistant?
Trust problem
Any observability tool that phones home to a third party is dead on arrival in enterprise. Data must stay in your infra.
Prometheus model for AI assistants
Crucible provides the collector. Company provides the destination. Zero data touches Crucible servers.
The closed feedback loop
Forge manages what configs/skills agents get. Ember measures how those configs/skills are actually used. Together: configure → measure → optimize.
Enterprise value
For Engineering Managers
"How much are we spending on AI per team? Which teams are actually using the tools we paid for? What's the ROI?"
For Platform Teams
"Which skills are used? Which should we deprecate? Are developers accepting or reverting AI suggestions?"
For Compliance & Security
"Audit trail of AI usage. Proof that guardrails are applied. Data stays in our infrastructure, zero third-party exposure."
Built on top of existing telemetry
Claude Code already ships with built-in OTEL instrumentation — skill activation events, token usage, session traces. Ember doesn't replace that. It builds on it.
Claude Code OTEL
Native claude_code.skill_activated events, token usage, session traces. One tool, one format.
+ Ember
Versioned skill metrics, cross-tool normalization (Cursor, Windsurf, Copilot too), unified dashboards, and flexible routing to any backend.
= Complete picture
One collector that ingests telemetry from all vendors, enriches it with version info and team context, and pushes to your infra.
Ember is not an alternative to vendor telemetry. It's the unified layer that makes all of it useful.
What you'll see
SKILL INVOCATIONS — LAST 7 DAYS
TOKEN USAGE BY TEAM
ASSISTANT DISTRIBUTION
SKILL VERSION ADOPTION
Mock data — illustrates what Ember dashboards will provide. Push to Grafana, DataDog, or any OTLP-compatible backend.
How Ember fits the landscape
| Tool | What it does | Ember's relationship |
|---|---|---|
| Claude Code OTEL | Native skill_activated events, token traces | Ember ingests this + adds versioning, cross-tool view |
| Langfuse | LLM app tracing (API calls) | Different scope — Langfuse tracks app LLM calls, Ember tracks developer AI usage |
| Helicone | LLM proxy with logging | Requires routing through their proxy. Enterprise won't do this. |
| DataDog LLM Obs | LLM monitoring | Ember can push to DataDog as a destination |
| OpenTelemetry | General observability standard | Protocol, not product. Ember exports OTLP natively. |
Ember doesn't compete with any of these. It's the unified collector that covers all AI assistants and pushes to your choice of backend.