Expand test coverage across server, agent, and web; fix bugs found during audit.

Adds hundreds of unit/integration/e2e tests, fixes WS bcrypt auth, config merge, fleet analytics, agent schedule/log tail, and documents stale PROBLEMS items. Updates PROBLEMS.md, README, and test scripts; ignores local spread-kits and coverage dirs.

agent/miner/pool_exports_test.go

@@ -0,0 +1,120 @@
+package miner
+
+import (
+	"strings"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	"crypto-miner-agent/config"
+	"crypto-miner-agent/job"
+	"crypto-miner-agent/stats"
+)
+
+func testPoolCfg() config.RuntimeConfig {
+	return config.RuntimeConfig{BuiltinConfig: config.BuiltinConfig{
+		MiningMode:   "always",
+		MaxCPUUsage:  0,
+		MaxMemoryPct: 0,
+		MinFreeRAM:   0,
+	}}
+}
+
+func TestNewPoolDefaultsThreads(t *testing.T) {
+	p := NewPool(0, testPoolCfg(), stats.NewReporter(), nil)
+	if p == nil || len(p.engines) != 1 {
+		t.Fatalf("threads<=0 should default to 1 engine, got %d", len(p.engines))
+	}
+}
+
+func TestNewPoolMultipleEngines(t *testing.T) {
+	p := NewPool(3, testPoolCfg(), stats.NewReporter(), nil)
+	if len(p.engines) != 3 {
+		t.Fatalf("expected 3 engines, got %d", len(p.engines))
+	}
+}
+
+func TestPoolSetJobNil(t *testing.T) {
+	p := NewPool(1, testPoolCfg(), stats.NewReporter(), nil)
+	p.SetJob(nil)
+	p.mu.RLock()
+	defer p.mu.RUnlock()
+	if p.currentJob != nil {
+		t.Fatal("SetJob(nil) should leave current job nil")
+	}
+}
+
+func TestPoolSetJobDerivesSeedFromBlob(t *testing.T) {
+	p := NewPool(1, testPoolCfg(), stats.NewReporter(), nil)
+	blobPrefix := strings.Repeat("ee", 32)
+	blob := blobPrefix + strings.Repeat("11", 22)
+	j := &job.Job{ID: "j1", Blob: blob}
+	p.SetJob(j)
+	if p.engines[0].seedHex != blobPrefix {
+		t.Fatalf("seed from blob prefix = %q, want %q", p.engines[0].seedHex, blobPrefix)
+	}
+}
+
+func TestPoolHashesPerSecondAndReset(t *testing.T) {
+	p := NewPool(1, testPoolCfg(), stats.NewReporter(), nil)
+	p.hashesTotal.Store(1000)
+	p.hashesLastReset = time.Now().Add(-2 * time.Second)
+	rate := p.HashesPerSecond()
+	if rate <= 0 {
+		t.Fatalf("expected positive hashrate, got %f", rate)
+	}
+	p.ResetHashCounter()
+	if p.hashesTotal.Load() != 0 {
+		t.Fatal("ResetHashCounter should zero hashesTotal")
+	}
+	if p.HashesPerSecond() != 0 {
+		t.Fatalf("hashrate after reset with no hashes should be 0, got %f", p.HashesPerSecond())
+	}
+}
+
+func TestPoolRemotePause(t *testing.T) {
+	p := NewPool(1, testPoolCfg(), stats.NewReporter(), nil)
+	if p.IsRemotePaused() {
+		t.Fatal("new pool should not be remote-paused")
+	}
+	p.PauseRemote()
+	if !p.IsRemotePaused() {
+		t.Fatal("PauseRemote should set flag")
+	}
+	p.ResumeRemote()
+	if p.IsRemotePaused() {
+		t.Fatal("ResumeRemote should clear flag")
+	}
+}
+
+func TestPoolSetShareHandler(t *testing.T) {
+	p := NewPool(1, testPoolCfg(), stats.NewReporter(), nil)
+	var called atomic.Bool
+	p.SetShareHandler(func(jobID, nonce, hash string) {
+		called.Store(true)
+	})
+	p.handlerMu.RLock()
+	h := p.handler
+	p.handlerMu.RUnlock()
+	if h == nil {
+		t.Fatal("handler should be set")
+	}
+	h("j", "n", "h")
+	if !called.Load() {
+		t.Fatal("swapped handler should run")
+	}
+}
+
+func TestPoolMiningAllowedAlways(t *testing.T) {
+	p := NewPool(1, testPoolCfg(), stats.NewReporter(), nil)
+	if !p.miningAllowed() {
+		t.Fatal("always mode with resource limits disabled should allow mining")
+	}
+}
+
+func TestPoolResourcesOKDisabledLimits(t *testing.T) {
+	p := NewPool(1, testPoolCfg(), stats.NewReporter(), nil)
+	if !p.resourcesOK() {
+		t.Fatal("zero resource limits should allow mining")
+	}
+}