fix: fix: add idle-after-final-message watchdog around claude -p to mitigate upstream Claude Code hang (#606)
All checks were successful
ci/woodpecker/push/ci Pipeline was successful
ci/woodpecker/pr/ci Pipeline was successful

This commit is contained in:
Claude 2026-04-10 17:33:56 +00:00
parent cd9937a4b4
commit ec5eb48224
2 changed files with 90 additions and 5 deletions

View file

@ -27,6 +27,88 @@ agent_recover_session() {
fi
}
# claude_run_with_watchdog — run claude with idle-after-final-message watchdog
#
# Mitigates upstream Claude Code hang (#591) by detecting when the final
# assistant message has been written and terminating the process after a
# short grace period instead of waiting for CLAUDE_TIMEOUT.
#
# The watchdog:
# 1. Streams claude stdout to a temp file
# 2. Polls for the final result marker ("type":"result" for stream-json
# or closing } for regular json output)
# 3. After detecting the final marker, starts a CLAUDE_IDLE_GRACE countdown
# 4. SIGTERM claude if it hasn't exited cleanly within the grace period
# 5. Falls back to CLAUDE_TIMEOUT as the absolute hard ceiling
#
# Usage: claude_run_with_watchdog claude [args...]
# Expects: LOGFILE, CLAUDE_TIMEOUT, CLAUDE_IDLE_GRACE (default 30)
# Returns: exit code from claude or timeout
claude_run_with_watchdog() {
local -a cmd=("$@")
local out_file pid grace_pid rc
# Create temp file for stdout capture
out_file=$(mktemp) || return 1
trap 'rm -f "$out_file"' RETURN
# Start claude in background, capturing stdout to temp file
"${cmd[@]}" > "$out_file" 2>>"$LOGFILE" &
pid=$!
# Background watchdog: poll for final result marker
(
local grace="${CLAUDE_IDLE_GRACE:-30}"
local detected=0
while kill -0 "$pid" 2>/dev/null; do
# Check for stream-json result marker first (more reliable)
if grep -q '"type":"result"' "$out_file" 2>/dev/null; then
detected=1
break
fi
# Fallback: check for closing brace of top-level result object
if tail -c 100 "$out_file" 2>/dev/null | grep -q '}[[:space:]]*$'; then
# Verify it looks like a JSON result (has session_id or result key)
if grep -qE '"(session_id|result)":' "$out_file" 2>/dev/null; then
detected=1
break
fi
fi
sleep 2
done
# If we detected a final message, wait grace period then kill if still running
if [ "$detected" -eq 1 ] && kill -0 "$pid" 2>/dev/null; then
log "watchdog: final result detected, ${grace}s grace period before SIGTERM"
sleep "$grace"
if kill -0 "$pid" 2>/dev/null; then
log "watchdog: claude -p idle for ${grace}s after final result; SIGTERM"
kill -TERM "$pid" 2>/dev/null || true
# Give it a moment to clean up
sleep 5
if kill -0 "$pid" 2>/dev/null; then
log "watchdog: force kill after SIGTERM timeout"
kill -KILL "$pid" 2>/dev/null || true
fi
fi
fi
) &
grace_pid=$!
# Hard ceiling timeout (existing behavior) — use tail --pid to wait for process
timeout --foreground "${CLAUDE_TIMEOUT:-7200}" tail --pid="$pid" -f /dev/null 2>/dev/null
rc=$?
# Clean up the watchdog
kill "$grace_pid" 2>/dev/null || true
wait "$grace_pid" 2>/dev/null || true
# Output the captured stdout
cat "$out_file"
return "$rc"
}
# agent_run — synchronous Claude invocation (one-shot claude -p)
# Usage: agent_run [--resume SESSION_ID] [--worktree DIR] PROMPT
# Sets: _AGENT_SESSION_ID (updated each call, persisted to SID_FILE)
@ -50,7 +132,7 @@ agent_run() {
mkdir -p "$(dirname "$lock_file")"
local output rc
log "agent_run: starting (resume=${resume_id:-(new)}, dir=${run_dir})"
output=$(cd "$run_dir" && flock -w 600 "$lock_file" timeout "${CLAUDE_TIMEOUT:-7200}" claude "${args[@]}" 2>>"$LOGFILE") && rc=0 || rc=$?
output=$(cd "$run_dir" && flock -w 600 "$lock_file" claude_run_with_watchdog claude "${args[@]}" 2>>"$LOGFILE") && rc=0 || rc=$?
if [ "$rc" -eq 124 ]; then
log "agent_run: timeout after ${CLAUDE_TIMEOUT:-7200}s (exit code $rc)"
elif [ "$rc" -ne 0 ]; then
@ -91,7 +173,7 @@ agent_run() {
local nudge="You stopped but did not push any code. You have uncommitted changes. Commit them and push."
log "agent_run: nudging (uncommitted changes)"
local nudge_rc
output=$(cd "$run_dir" && flock -w 600 "$lock_file" timeout "${CLAUDE_TIMEOUT:-7200}" claude -p "$nudge" --resume "$_AGENT_SESSION_ID" --output-format json --dangerously-skip-permissions --max-turns 50 ${CLAUDE_MODEL:+--model "$CLAUDE_MODEL"} 2>>"$LOGFILE") && nudge_rc=0 || nudge_rc=$?
output=$(cd "$run_dir" && flock -w 600 "$lock_file" claude_run_with_watchdog claude -p "$nudge" --resume "$_AGENT_SESSION_ID" --output-format json --dangerously-skip-permissions --max-turns 50 ${CLAUDE_MODEL:+--model "$CLAUDE_MODEL"} 2>>"$LOGFILE") && nudge_rc=0 || nudge_rc=$?
if [ "$nudge_rc" -eq 124 ]; then
log "agent_run: nudge timeout after ${CLAUDE_TIMEOUT:-7200}s (exit code $nudge_rc)"
elif [ "$nudge_rc" -ne 0 ]; then

View file

@ -28,7 +28,10 @@
# ops_commit_and_push MESSAGE [FILES] — commit/push to ops repo
# cleanup_stale_crashed_worktrees [HOURS] — thin wrapper around worktree_cleanup_stale
#
# Requires: lib/env.sh, lib/worktree.sh sourced first for shared helpers.
# Requires: lib/env.sh, lib/worktree.sh, lib/agent-sdk.sh sourced first for shared helpers.
# Source agent-sdk for claude_run_with_watchdog watchdog helper
source "$(dirname "$0")/agent-sdk.sh"
# ── Run guards ───────────────────────────────────────────────────────────
@ -248,7 +251,7 @@ Write the complete, rewritten lessons-learned.md content below. No preamble, no
# Run claude -p one-shot with same model as agent
local output
output=$(claude -p "$digest_prompt" \
output=$(claude_run_with_watchdog claude -p "$digest_prompt" \
--output-format json \
--dangerously-skip-permissions \
${model:+--model "$model"} \
@ -442,7 +445,7 @@ Write the journal entry below. Use markdown format."
# Run claude -p one-shot with same model as agent
local output
output=$(claude -p "$reflection_prompt" \
output=$(claude_run_with_watchdog claude -p "$reflection_prompt" \
--output-format json \
--dangerously-skip-permissions \
${CLAUDE_MODEL:+--model "$CLAUDE_MODEL"} \