From cf83444a9d341ae362e65a9f995100c69176887c Mon Sep 17 00:00:00 2001
From: Peter Stone <thepeterstone@gmail.com>
Date: Thu, 5 Mar 2026 18:51:50 +0000
Subject: Rescue work from claudomator-work: question/answer, ratelimit,
 start-next-task

Merges features developed in /site/doot.terst.org/claudomator-work (a
stale clone) into the canonical repo:

- executor: QuestionRegistry for human-in-the-loop answers, rate limit
  detection and exponential backoff retry (ratelimit.go, question.go)
- executor/claude.go: process group isolation (SIGKILL orphans on cancel),
  os.Pipe for reliable stdout drain, backoff retry on rate limits
- api/scripts.go: POST /api/scripts/start-next-task handler
- api/server.go: startNextTaskScript field, answer-question route,
  BroadcastQuestion for WebSocket question events
- web: Cancel/Restart buttons, question banner UI, log viewer, validate
  section, WebSocket auto-connect

All tests pass.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---
 internal/executor/ratelimit.go | 76 ++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 76 insertions(+)
 create mode 100644 internal/executor/ratelimit.go

(limited to 'internal/executor/ratelimit.go')

diff --git a/internal/executor/ratelimit.go b/internal/executor/ratelimit.go
new file mode 100644
index 0000000..884da43
--- /dev/null
+++ b/internal/executor/ratelimit.go
@@ -0,0 +1,76 @@
+package executor
+
+import (
+	"context"
+	"fmt"
+	"regexp"
+	"strconv"
+	"strings"
+	"time"
+)
+
+var retryAfterRe = regexp.MustCompile(`(?i)retry[-_ ]after[:\s]+(\d+)`)
+
+const maxBackoffDelay = 5 * time.Minute
+
+// isRateLimitError returns true if err looks like a Claude API rate-limit response.
+func isRateLimitError(err error) bool {
+	if err == nil {
+		return false
+	}
+	msg := strings.ToLower(err.Error())
+	return strings.Contains(msg, "rate limit") ||
+		strings.Contains(msg, "too many requests") ||
+		strings.Contains(msg, "429") ||
+		strings.Contains(msg, "overloaded")
+}
+
+// parseRetryAfter extracts a Retry-After duration from an error message.
+// Returns 0 if no retry-after value is found.
+func parseRetryAfter(msg string) time.Duration {
+	m := retryAfterRe.FindStringSubmatch(msg)
+	if m == nil {
+		return 0
+	}
+	secs, err := strconv.Atoi(m[1])
+	if err != nil || secs <= 0 {
+		return 0
+	}
+	return time.Duration(secs) * time.Second
+}
+
+// runWithBackoff calls fn repeatedly on rate-limit errors, using exponential backoff.
+// maxRetries is the max number of retry attempts (not counting the initial call).
+// baseDelay is the initial backoff duration (doubled each retry).
+func runWithBackoff(ctx context.Context, maxRetries int, baseDelay time.Duration, fn func() error) error {
+	var lastErr error
+	for attempt := 0; attempt <= maxRetries; attempt++ {
+		lastErr = fn()
+		if lastErr == nil {
+			return nil
+		}
+		if !isRateLimitError(lastErr) {
+			return lastErr
+		}
+		if attempt == maxRetries {
+			break
+		}
+
+		// Compute exponential backoff delay.
+		delay := baseDelay * (1 << attempt)
+		if delay > maxBackoffDelay {
+			delay = maxBackoffDelay
+		}
+		// Use Retry-After header value if present.
+		if ra := parseRetryAfter(lastErr.Error()); ra > 0 {
+			delay = ra
+		}
+
+		select {
+		case <-ctx.Done():
+			return fmt.Errorf("context cancelled during rate-limit backoff: %w", ctx.Err())
+		case <-time.After(delay):
+		}
+	}
+	return lastErr
+}
-- 
cgit v1.2.3