* fix(telemetry): report token usage for SSE streaming paths
ParseSSEStream already parsed the usage block from SSE chunks but only
printed it, so claw402 streaming calls (and native streaming) never
fired TokenUsageCallback. GA4 therefore undercounted AI usage on the
streaming path.
Return the parsed usage from ParseSSEStream and have both callers fire
the callback with their own Provider/Model.
* chore: drop leftover debug Printf in ParseSSEStream
Telemetry is now wired via TokenUsageCallback, so the Printf is
redundant noise in the stream path.
When CLAW402_WALLET_KEY env var is set, all nofxos.ai data API calls
(AI500, OI rankings, NetFlow, price rankings) are automatically routed
through claw402.ai with x402 USDC micropayment.
- provider/nofxos/claw402.go: x402 GET request client for data APIs
- provider/nofxos/client.go: claw402 mode support in doRequest()
- kernel/engine.go: auto-detect CLAW402_WALLET_KEY and enable routing
- mcp/payment/x402.go: MakeClaw402SignFunc helper
Without CLAW402_WALLET_KEY, falls back to direct nofxos.ai (backward compat).
- Extract ParseSSEStream as shared function from CallWithRequestStream
- Add DoX402RequestStream and X402CallStream for streaming x402 payments
- Switch Claw402Client.Call to use streaming (X402CallStream)
- TeeReader fallback: SSE parsing with JSON fallback for non-SSE responses
- Idle timeout watchdog (90s) protects against stalled streams
AI inference (especially DeepSeek) often exceeds the default 120s HTTP
timeout, causing the client to disconnect while the server completes
successfully — resulting in repeated payments on each retry.
Changes:
- Set X402Timeout = 5min for all x402 providers (Claw402, BlockRunBase, BlockRunSol)
- Handle 402 during payment retry by re-extracting Payment-Required
header and re-signing instead of failing immediately
- Increase payment retry attempts from 3 to 5 for unstable gateways
