feat(providers): success-rate accounting

Track per-provider and per-(provider,upstream-model) upstream-attempt
outcomes using the existing fixed-window counter machinery (limits.go),
following mcp_a2a.go's per-target counter pattern: kind-embedded keys
(kindProvider="prov", kindProviderModel="provmodel") keep a provider
scope collision-safe against a user/group/MCP-target/agent of the same
name.

limiter.recordProviderAttempt(provider, model, resp, err) increments an
"attempts" counter on every call, and a "failures" counter only for a
provider-fault outcome — reusing retry.go's own isTransient(resp, err)
classification exactly (transport/connection error, HTTP 429, or a 5xx),
not forking it. Both counters are written at minute and day granularity
only; provider health is a now-and-today question, not a billing one, so
there is no month window. limiter.providerUsage batches every scope's
read into one storeGetMulti round trip, mirroring targetUsage's own
single-batch discipline.

Wired at every chokepoint an upstream HTTP attempt resolves:
- providers.go's upstreamJSON/upstreamRawBytes path (chat, embeddings,
  images, audio) via retry.go's retryPolicy.do, which now looks up an
  optional attemptRecorder from ctx and invokes it once per retry
  attempt (attempt-accounting: a request retried twice before
  succeeding reports three attempts, not one).
- routes_unified.go's runUnified (chat/embeddings, streaming and
  non-streaming alike — both funnel through the same upstreamJSON call).
- routes_media.go's three media handlers (images.generations,
  audio.speech, audio.transcriptions).
- routes_passthrough.go's proxyUpstream (native provider passthrough) —
  a single un-retried attempt, provider-level only: the upstream model
  lives in the response body, read only after the attempt resolves.

The attemptRecorder is threaded via context (providers.go's
withAttemptRecorder/attemptRecorderFromContext), not a new function
parameter on every call site — proxyUpstream and retryPolicy.do are
both shared with paths that must NOT record provider accounting
(registry.go's discovery listModels calls, and the MCP/A2A target
proxy, mcp_a2a.go's handleTargetProxy), so scoping via context keeps
their signatures unchanged and those callers correctly opt out by
simply never wrapping their context.

retry.go

@@ -215,13 +215,26 @@ func drainAndClose(resp *http.Response) {
 // last attempt's (resp, err) as-is either way: on success, on a
 // still-transient failure once tries are exhausted, on a non-transient
 // failure at any attempt, or on a wait ctx aborted early.
+//
+// Feature A (v0.22) attempt-accounting: every call() invocation — not just
+// the final one — is reported to ctx's attemptRecorder, if it carries one
+// (attemptRecorderFromContext, providers.go), before the transient check
+// above decides whether to retry. A request retried twice before
+// succeeding therefore reports THREE attempts, not one — the caller
+// (limiter.recordProviderAttempt, limits.go) classifies each with the
+// identical isTransient(resp, err) this loop already uses, so "attempt"
+// and "failure" here always agree with what actually got retried.
 func (p *retryPolicy) do(ctx context.Context, call func() (*http.Response, error)) (*http.Response, error) {
 	tries := p.totalTries()
+	rec := attemptRecorderFromContext(ctx)
 
 	var resp *http.Response
 	var err error
 	for attempt := 1; attempt <= tries; attempt++ {
 		resp, err = call()
+		if rec != nil {
+			rec(resp, err)
+		}
 		if !isTransient(resp, err) || attempt == tries {
 			return resp, err
 		}