fixup! chore: update marketplace for v0.11.37

march-improvements

pkg/hooks/response.go

@@ -2,6 +2,7 @@
 package hooks
 
 import (
+	"context"
 	"crypto/sha256"
 	"encoding/hex"
 	"encoding/json"
@@ -9,6 +10,7 @@ import (
 	"io"
 	"os"
 	"path/filepath"
+	"time"
 )
 
 // HookResponse is the response sent back to Claude Code.
@@ -31,6 +33,14 @@ func ProjectIDWithName(cwd string) string {
 	return fmt.Sprintf("%s_%s", dirName, shortHash)
 }
 
+// HookDeadline returns a context with the hook's timeout budget minus a safety margin.
+// This ensures hooks return gracefully before Claude kills them.
+func HookDeadline(timeout time.Duration) (context.Context, context.CancelFunc) {
+	// Use 80% of the timeout to leave margin for response serialization
+	safeTimeout := time.Duration(float64(timeout) * 0.8)
+	return context.WithTimeout(context.Background(), safeTimeout)
+}
+
 // Exit codes for Claude Code hooks
 const (
 	ExitSuccess         = 0
@@ -92,7 +102,7 @@ func RunHook[T any](hookName string, handler HookHandler[T]) {
 
 	// Parse input
 	var input T
-	if err := json.Unmarshal(inputData, &input); err != nil {
+	if err = json.Unmarshal(inputData, &input); err != nil {
 		WriteError(hookName, err)
 		os.Exit(1)
 	}

pkg/hooks/worker.go

@@ -3,6 +3,7 @@ package hooks
 
 import (
 	"bytes"
+	"context"
 	"encoding/json"
 	"fmt"
 	"net"
@@ -12,6 +13,8 @@ import (
 	"path/filepath"
 	"strconv"
 	"strings"
+	"sync"
+	"syscall"
 	"time"
 )
 
@@ -22,13 +25,53 @@ const (
 	// DefaultWorkerPort is the default worker port.
 	DefaultWorkerPort = 37777
 
-	// HealthCheckTimeout is the timeout for health checks (reduced from 5s for faster startup).
-	HealthCheckTimeout = 1 * time.Second
+	// HealthCheckTimeout is the timeout for health checks.
+	HealthCheckTimeout = 2 * time.Second
 
 	// StartupTimeout is the timeout for worker startup.
 	StartupTimeout = 30 * time.Second
+
+	// workerCacheMaxAge is how long the worker cache is considered fresh.
+	workerCacheMaxAge = 10 * time.Second
+
+	// circuitBreakerCooldown is how long to wait after a startup failure before retrying.
+	circuitBreakerCooldown = 30 * time.Second
+
+	// healthCheckRetries is the number of health check attempts before declaring dead.
+	healthCheckRetries = 3
+
+	// healthCheckRetryDelay is the delay between health check retries.
+	healthCheckRetryDelay = 200 * time.Millisecond
 )
 
+var (
+	// circuitBreakerMu protects lastStartupFailure.
+	circuitBreakerMu   sync.Mutex
+	lastStartupFailure time.Time
+)
+
+// IsWorkerAvailable performs a fast check without network calls.
+// Returns true if the worker is likely available, false if definitely down.
+func IsWorkerAvailable() bool {
+	// Check circuit breaker first
+	circuitBreakerMu.Lock()
+	if !lastStartupFailure.IsZero() && time.Since(lastStartupFailure) < circuitBreakerCooldown {
+		circuitBreakerMu.Unlock()
+		return false
+	}
+	circuitBreakerMu.Unlock()
+
+	// Check PID cache
+	entry := readWorkerCache()
+	if entry == nil {
+		return true // No cache = unknown, don't block
+	}
+
+	// Cache exists and is fresh (readWorkerCache already checks staleness)
+	// Check if cached process is alive
+	return isProcessAlive(entry.PID)
+}
+
 // GetWorkerPort returns the worker port from environment or default.
 func GetWorkerPort() int {
 	if port := os.Getenv("CLAUDE_MNEMONIC_WORKER_PORT"); port != "" {
@@ -40,29 +83,149 @@ func GetWorkerPort() int {
 }
 
 // IsWorkerRunning checks if the worker is running and healthy.
+// Parses the JSON health response to check the "ready" field when available.
+// Falls back to HTTP status code check for backwards compatibility.
 func IsWorkerRunning(port int) bool {
 	client := &http.Client{Timeout: HealthCheckTimeout}
 	resp, err := client.Get(fmt.Sprintf("http://127.0.0.1:%d/api/health", port))
 	if err != nil {
 		return false
 	}
-	defer resp.Body.Close()
+	defer func() { _ = resp.Body.Close() }()
+
+	// Try to parse JSON response for structured health check
+	var health struct {
+		Ready bool `json:"ready"`
+	}
+	if err := json.NewDecoder(resp.Body).Decode(&health); err == nil {
+		return health.Ready
+	}
+
+	// Fallback: treat HTTP 200 as healthy (backwards compatibility)
 	return resp.StatusCode == http.StatusOK
 }
 
+// workerCachePath returns the path to the worker cache file.
+func workerCachePath() string {
+	home := os.Getenv("HOME")
+	if home == "" {
+		return ""
+	}
+	return filepath.Join(home, ".claude-mnemonic", ".worker-cache")
+}
+
+// workerCacheEntry holds cached worker state: "port:pid:timestamp".
+type workerCacheEntry struct {
+	Timestamp time.Time
+	Port      int
+	PID       int
+}
+
+// readWorkerCache reads the worker cache file and returns the entry if fresh.
+func readWorkerCache() *workerCacheEntry {
+	path := workerCachePath()
+	if path == "" {
+		return nil
+	}
+	data, err := os.ReadFile(path)
+	if err != nil {
+		return nil
+	}
+	parts := strings.SplitN(strings.TrimSpace(string(data)), ":", 3)
+	if len(parts) != 3 {
+		return nil
+	}
+	port, err := strconv.Atoi(parts[0])
+	if err != nil || port <= 0 {
+		return nil
+	}
+	pid, err := strconv.Atoi(parts[1])
+	if err != nil || pid <= 0 {
+		return nil
+	}
+	ts, err := strconv.ParseInt(parts[2], 10, 64)
+	if err != nil {
+		return nil
+	}
+	entry := &workerCacheEntry{
+		Port:      port,
+		PID:       pid,
+		Timestamp: time.Unix(ts, 0),
+	}
+	// Check freshness
+	if time.Since(entry.Timestamp) > workerCacheMaxAge {
+		return nil
+	}
+	return entry
+}
+
+// writeWorkerCache writes the worker cache file.
+func writeWorkerCache(port, pid int) {
+	path := workerCachePath()
+	if path == "" {
+		return
+	}
+	// Ensure directory exists
+	dir := filepath.Dir(path)
+	_ = os.MkdirAll(dir, 0o700)
+	data := fmt.Sprintf("%d:%d:%d", port, pid, time.Now().Unix())
+	_ = os.WriteFile(path, []byte(data), 0o600)
+}
+
+// isProcessAlive checks if a process with the given PID exists and is alive.
+func isProcessAlive(pid int) bool {
+	proc, err := os.FindProcess(pid)
+	if err != nil {
+		return false
+	}
+	// Signal 0 checks if process exists without actually sending a signal.
+	err = proc.Signal(syscall.Signal(0))
+	return err == nil
+}
+
+// isWorkerRunningWithRetries checks if the worker is running, retrying on timeout.
+// Returns true only if health check succeeds. Returns false if all retries fail.
+func isWorkerRunningWithRetries(port int) bool {
+	for i := 0; i < healthCheckRetries; i++ {
+		if IsWorkerRunning(port) {
+			return true
+		}
+		if i < healthCheckRetries-1 {
+			time.Sleep(healthCheckRetryDelay)
+		}
+	}
+	return false
+}
+
 // EnsureWorkerRunning ensures the worker is running, starting it if necessary.
 // If a worker is already running and healthy with matching version, it reuses it.
 // If version mismatch or unhealthy, it kills the old worker and starts fresh.
 func EnsureWorkerRunning() (int, error) {
 	port := GetWorkerPort()
 
-	// Check if already running and healthy
-	if IsWorkerRunning(port) {
+	// Fast path: check PID cache before making any HTTP calls.
+	if entry := readWorkerCache(); entry != nil && entry.Port == port {
+		if isProcessAlive(entry.PID) {
+			return port, nil
+		}
+	}
+
+	// Circuit breaker: if we failed to start recently, don't retry immediately.
+	circuitBreakerMu.Lock()
+	if !lastStartupFailure.IsZero() && time.Since(lastStartupFailure) < circuitBreakerCooldown {
+		circuitBreakerMu.Unlock()
+		return 0, fmt.Errorf("worker startup failed recently (circuit breaker open, retry after %v)", circuitBreakerCooldown-time.Since(lastStartupFailure))
+	}
+	circuitBreakerMu.Unlock()
+
+	// Check if already running and healthy (with retries to avoid false negatives under load)
+	if isWorkerRunningWithRetries(port) {
 		// Check version - if mismatch, restart (unless both are dev builds)
 		if runningVersion := GetWorkerVersion(port); runningVersion != "" {
 			if runningVersion != Version {
 				// For dev/dirty builds, don't restart if base versions match
 				if versionsCompatible(runningVersion, Version) {
+					updateCacheFromPort(port)
 					return port, nil
 				}
 				fmt.Fprintf(os.Stderr, "[claude-mnemonic] Worker version mismatch (running: %s, expected: %s), restarting...\n", runningVersion, Version)
@@ -72,23 +235,34 @@ func EnsureWorkerRunning() (int, error) {
 				time.Sleep(500 * time.Millisecond)
 			} else {
 				// Version matches, reuse existing worker
+				updateCacheFromPort(port)
 				return port, nil
 			}
 		} else {
 			// Couldn't get version, assume it's fine
+			updateCacheFromPort(port)
 			return port, nil
 		}
 	}
 
-	// Check if port is in use but worker is unhealthy
+	// Port is in use but health check failed -- worker may be slow, not dead.
 	if IsPortInUse(port) {
-		// Something is using the port but not responding to health checks
-		// Try to kill it
+		// The port is responding to TCP but health check timed out.
+		// Don't kill it -- it's likely just under load. Give it more time.
+		fmt.Fprintf(os.Stderr, "[claude-mnemonic] Worker on port %d is slow to respond, waiting...\n", port)
+		// Try a few more times with longer delays before giving up
+		for i := 0; i < 3; i++ {
+			time.Sleep(500 * time.Millisecond)
+			if IsWorkerRunning(port) {
+				updateCacheFromPort(port)
+				return port, nil
+			}
+		}
+		// Still not healthy after extended wait -- kill and restart
+		fmt.Fprintf(os.Stderr, "[claude-mnemonic] Worker unresponsive after extended wait, restarting...\n")
 		if err := KillProcessOnPort(port); err != nil {
-			// Log but continue - maybe it will die on its own
 			fmt.Fprintf(os.Stderr, "[claude-mnemonic] Warning: failed to kill unhealthy process on port %d: %v\n", port, err)
 		}
-		// Wait a moment for port to be released
 		time.Sleep(500 * time.Millisecond)
 	}
 
@@ -103,9 +277,14 @@ func EnsureWorkerRunning() (int, error) {
 	cmd.Stdout = os.Stderr
 	cmd.Stderr = os.Stderr
 	if err := cmd.Start(); err != nil {
+		circuitBreakerMu.Lock()
+		lastStartupFailure = time.Now()
+		circuitBreakerMu.Unlock()
 		return 0, fmt.Errorf("failed to start worker: %w", err)
 	}
 
+	pid := cmd.Process.Pid
+
 	// Wait for worker to be ready with exponential backoff
 	deadline := time.Now().Add(StartupTimeout)
 	backoff := 50 * time.Millisecond
@@ -113,6 +292,7 @@ func EnsureWorkerRunning() (int, error) {
 
 	for time.Now().Before(deadline) {
 		if IsWorkerRunning(port) {
+			writeWorkerCache(port, pid)
 			return port, nil
 		}
 		time.Sleep(backoff)
@@ -123,9 +303,31 @@ func EnsureWorkerRunning() (int, error) {
 		}
 	}
 
+	circuitBreakerMu.Lock()
+	lastStartupFailure = time.Now()
+	circuitBreakerMu.Unlock()
 	return 0, fmt.Errorf("worker failed to start within timeout")
 }
 
+// updateCacheFromPort finds the PID of the process on the port and updates the cache.
+func updateCacheFromPort(port int) {
+	cmd := exec.Command("lsof", "-t", "-i", fmt.Sprintf(":%d", port)) // #nosec G204 -- port is from internal config
+	output, err := cmd.Output()
+	if err != nil {
+		return
+	}
+	pidStr := strings.TrimSpace(string(output))
+	// Take first PID if multiple
+	if idx := strings.Index(pidStr, "\n"); idx > 0 {
+		pidStr = pidStr[:idx]
+	}
+	pid, err := strconv.Atoi(pidStr)
+	if err != nil || pid <= 0 {
+		return
+	}
+	writeWorkerCache(port, pid)
+}
+
 // GetWorkerVersion gets the version of the running worker.
 func GetWorkerVersion(port int) string {
 	client := &http.Client{Timeout: HealthCheckTimeout}
@@ -133,7 +335,7 @@ func GetWorkerVersion(port int) string {
 	if err != nil {
 		return ""
 	}
-	defer resp.Body.Close()
+	defer func() { _ = resp.Body.Close() }()
 
 	if resp.StatusCode != http.StatusOK {
 		return ""
@@ -243,7 +445,7 @@ func POST(port int, path string, body interface{}) (map[string]interface{}, erro
 	if err != nil {
 		return nil, err
 	}
-	defer resp.Body.Close()
+	defer func() { _ = resp.Body.Close() }()
 
 	if resp.StatusCode >= 400 {
 		return nil, fmt.Errorf("request failed: %s", resp.Status)
@@ -251,13 +453,38 @@ func POST(port int, path string, body interface{}) (map[string]interface{}, erro
 
 	var result map[string]interface{}
 	if err := json.NewDecoder(resp.Body).Decode(&result); err != nil {
-		// Not all endpoints return JSON body - return empty map for success with no body
-		return map[string]interface{}{}, nil
+		// Not all endpoints return JSON
+		return nil, nil
 	}
 
 	return result, nil
 }
 
+// POSTWithContext sends a POST request using the provided context.
+// Used for fire-and-forget calls where we want to control the timeout externally.
+func POSTWithContext(ctx context.Context, port int, path string, body interface{}) error {
+	jsonBody, err := json.Marshal(body)
+	if err != nil {
+		return err
+	}
+
+	req, err := http.NewRequestWithContext(ctx, http.MethodPost,
+		fmt.Sprintf("http://127.0.0.1:%d%s", port, path),
+		bytes.NewReader(jsonBody))
+	if err != nil {
+		return err
+	}
+	req.Header.Set("Content-Type", "application/json")
+
+	client := &http.Client{Timeout: 10 * time.Second}
+	resp, err := client.Do(req)
+	if err != nil {
+		return err
+	}
+	defer func() { _ = resp.Body.Close() }()
+	return nil
+}
+
 // GET sends a GET request to the worker.
 func GET(port int, path string) (map[string]interface{}, error) {
 	client := &http.Client{Timeout: 10 * time.Second}
@@ -266,7 +493,7 @@ func GET(port int, path string) (map[string]interface{}, error) {
 	if err != nil {
 		return nil, err
 	}
-	defer resp.Body.Close()
+	defer func() { _ = resp.Body.Close() }()
 
 	if resp.StatusCode >= 400 {
 		return nil, fmt.Errorf("request failed: %s", resp.Status)

pkg/hooks/worker_test.go

@@ -517,7 +517,7 @@ func TestProjectIDWithName_Uniqueness(t *testing.T) {
 // TestHookConstants tests hook-related constants.
 func TestHookConstants(t *testing.T) {
 	assert.Equal(t, 37777, DefaultWorkerPort)
-	assert.Equal(t, 1*time.Second, HealthCheckTimeout)
+	assert.Equal(t, 2*time.Second, HealthCheckTimeout)
 	assert.Equal(t, 30*time.Second, StartupTimeout)
 }