contextpulse
Full-Stack MCP Observability

End-to-end tracing for agent sessions

Correlate transport handshakes, tool executions, parameter validations, and downstream latency into single distributed traces across Claude, Cursor, and custom agent runtimes.

Real-time Telemetry Pipeline

How ContextPulse Observes Every Protocol Frame

Zero-overhead async instrumentation at the JSON-RPC boundary.

1. AI Clients
Claude Desktop
protocol: stdio / sse
Cursor AI
protocol: stdio / sse
ChatGPT Operator
protocol: stdio / sse
ContextPulse SDKFail-Open BoundaryRedacts credentials & hashes payloads in-memory
0 prompts leaked
2. MCP Tools Executed
search_docs
latency: 48ms
Healthy
query_database
latency: 142ms
Healthy
send_email
latency: 890ms
Alert
Ingest: 1,240 events/sec
Avg overhead: < 0.4ms
100% Non-blocking Async

p95 Latency Profiles

Tail latency quietly wrecks multi-turn agent sessions. Track execution time by tool, isolating slow database queries from LLM network overhead.

Session Flow & Retries

When an agent calls a tool 4 times in a row, it’s stuck in a loop. ContextPulse flags loops instantly and estimates token waste.

Safe Bounded Capture

Collect diagnostic metadata without violating user confidentiality. Payload scrubbing removes bearer tokens and PII before transmission.