omnilane 0.20.0 → 0.31.0

package/scripts/dispatch.sh

@@ -3,9 +3,10 @@ set -euo pipefail
 # omnilane dispatch — one routing table, any harness.
 #
 # Usage:
-#   dispatch.sh [--background] [--dry-run] [--mode advise|work|sysops] [--workdir DIR]
+#   dispatch.sh [--background] [--live|--single-shot] [--dry-run]
+#               [--mode advise|work|sysops] [--workdir DIR]
 #               [--vendor V] [--model M] [--effort E] [--timeout SECONDS]
-#               [--job-timeout SECONDS] LANE "TASK TEXT"
+#               [--job-timeout SECONDS] [--idle-timeout SECONDS] LANE "TASK TEXT"
 #   dispatch.sh [--json] --list [--json]
 #   dispatch.sh [--json] --explain LANE [--json]
 #   dispatch.sh [--json] --validate [--json]
@@ -27,10 +28,12 @@ set -euo pipefail
 # A separate --job-timeout can cap lock wait plus all calls in this dispatch.
 
 source "$(dirname "${BASH_SOURCE[0]}")/lib/common.sh"
+# shellcheck disable=SC1091
+source "$OMNILANE_REPO/scripts/lib/live-protocol.sh"
 
 MODE="advise"; WORKDIR="$PWD"; BACKGROUND=0; DRY_RUN=0
 OVERRIDE_VENDOR=""; OVERRIDE_MODEL=""; OVERRIDE_EFFORT=""; OVERRIDE_TIMEOUT=""
-OVERRIDE_JOB_TIMEOUT=""
+OVERRIDE_JOB_TIMEOUT=""; OVERRIDE_IDLE_TIMEOUT=""; SESSION_REQUEST="auto"
 
 usage_error() {
   echo 'usage: dispatch.sh [--background] [--dry-run] [flags] LANE "TASK" | [--json] --list|--validate [--json] | [--json] --explain LANE [--json] | --help' >&2
@@ -47,6 +50,9 @@ Dispatch one task to the first available vendor CLI in LANE's fallback chain.
 A TASK of "-" reads the task text from stdin.
 
 flags:
+  --live                    require a resident session (background only)
+  --single-shot             force one-shot dispatch (background only)
+  --idle-timeout SECONDS    close an idle live mailbox (default 900; 0 disables)
   --background           run in the background and print the JOB_ID
   --dry-run              print the fully resolved dispatch plan and stop
                          before any provider call or job state
@@ -139,6 +145,8 @@ print_dry_run_plan() {
   print_dry_run_value workdir "$WORKDIR"
   printf 'timeout=%s\n' "$TIMEOUT"
   printf 'job_timeout=%s\n' "$job_timeout"
+  printf 'idle_timeout=%s\n' "$IDLE_TIMEOUT"
+  print_dry_run_value session_mode "$SESSION_MODE"
   printf 'candidate=%s/%s\n' "$RESOLVED_IDX" "$RESOLVED_TOTAL"
   printf 'background=%s\n' "$background"
   printf 'task_source=%s\n' "$task_source"
@@ -461,8 +469,18 @@ while [[ $# -gt 0 ]]; do
   case "$1" in
     --list|--explain|--validate|--json) usage_error ;;
     --background) BACKGROUND=1; shift ;;
+    --live)
+      [[ "$SESSION_REQUEST" != "single-shot" ]] || {
+        echo "omnilane: --live and --single-shot are mutually exclusive" >&2; exit 2
+      }
+      SESSION_REQUEST="live"; shift ;;
+    --single-shot)
+      [[ "$SESSION_REQUEST" != "live" ]] || {
+        echo "omnilane: --live and --single-shot are mutually exclusive" >&2; exit 2
+      }
+      SESSION_REQUEST="single-shot"; shift ;;
     --dry-run) DRY_RUN=1; shift ;;
-    --mode|--workdir|--vendor|--model|--effort|--timeout|--job-timeout)
+    --mode|--workdir|--vendor|--model|--effort|--timeout|--job-timeout|--idle-timeout)
       # Value-taking flags: a missing value must be a clean usage error (exit 2),
       # not a `set -u` "unbound variable" crash on $2.
       [[ $# -ge 2 ]] || { echo "omnilane: $1 needs a value" >&2; exit 2; }
@@ -474,6 +492,7 @@ while [[ $# -gt 0 ]]; do
         --effort) OVERRIDE_EFFORT="$2" ;;
         --timeout) OVERRIDE_TIMEOUT="$2" ;;
         --job-timeout) OVERRIDE_JOB_TIMEOUT="$2" ;;
+        --idle-timeout) OVERRIDE_IDLE_TIMEOUT="$2" ;;
       esac
       shift 2 ;;
     -*) echo "omnilane: unknown flag" >&2; exit 2 ;;
@@ -487,6 +506,11 @@ done
   echo 'omnilane: unexpected extra arguments; quote a multiword task' >&2
   exit 2
 }
+if [[ "$SESSION_REQUEST" != "auto" && "$BACKGROUND" -ne 1 ]]; then
+  echo "omnilane: --${SESSION_REQUEST} requires --background" >&2
+  exit 2
+fi
+
 LANE="$1"
 TASK="$2"
 [[ "$LANE" =~ ^[a-z][a-z0-9-]*$ ]] || { echo "omnilane: invalid lane name" >&2; exit 2; }
@@ -541,6 +565,21 @@ fi
 RUNNER="$OMNILANE_REPO/scripts/runners/run-$VENDOR.sh"
 [[ -x "$RUNNER" ]] || { echo "omnilane: no runner for vendor '$VENDOR'" >&2; exit 2; }
 
+SESSION_MODE="single-shot"
+if [[ "$SESSION_REQUEST" != "single-shot" ]] && live_vendor_capable "$VENDOR"; then
+  SESSION_MODE="live"
+fi
+if [[ "$SESSION_REQUEST" == "live" ]] && ! live_vendor_capable "$VENDOR"; then
+  echo "omnilane: vendor '$VENDOR' cannot run live; live-capable vendors: $(live_capable_vendors)" >&2
+  exit 2
+fi
+export OMNILANE_SESSION_MODE="$SESSION_MODE"
+if [[ "$SESSION_REQUEST" == "live" ]]; then
+  export OMNILANE_LIVE_REQUIRED=1
+else
+  export OMNILANE_LIVE_REQUIRED=0
+fi
+
 # Watchdog seconds: --timeout > per-lane OMNILANE_TIMEOUT_<LANE> > OMNILANE_TIMEOUT > 600.
 # Resolve here and export so every runner (and vote's sub-runners) inherits the
 # same value without a per-runner code change; they already read OMNILANE_TIMEOUT.
@@ -556,6 +595,14 @@ fi
 }
 export OMNILANE_TIMEOUT="$TIMEOUT"
 
+IDLE_TIMEOUT="$OVERRIDE_IDLE_TIMEOUT"
+[[ -n "$IDLE_TIMEOUT" ]] || IDLE_TIMEOUT="${OMNILANE_IDLE_TIMEOUT:-900}"
+[[ "$IDLE_TIMEOUT" =~ ^(0|[1-9][0-9]*)$ ]] || {
+  echo "omnilane: invalid idle timeout (want a non-negative integer of seconds)" >&2
+  exit 2
+}
+export OMNILANE_IDLE_TIMEOUT="$IDLE_TIMEOUT"
+
 JOB_SUPERVISOR="$OMNILANE_REPO/scripts/lib/job-timeout.pl"
 JOB_WORKER="$OMNILANE_REPO/scripts/lib/job-worker.sh"
 
@@ -649,9 +696,10 @@ else
 fi
 
 # meta "timeout" is the resolved per-CLI-call watchdog cap, not a whole-job total.
-(umask 077; printf '{"lane":"%s","vendor":"%s","model":"%s","effort":"%s","timeout":%s,"job_timeout":%s,"mode":"%s","workdir":"%s","foreman_session":"%s","candidate":"%s/%s","started":"%s"}\n' \
-  "$(json_escape "$LANE")" "$(json_escape "$VENDOR")" "$(json_escape "$MODEL")" \
-  "$(json_escape "$EFFORT")" "$TIMEOUT" "$JOB_TIMEOUT_JSON" "$(json_escape "$MODE")" \
+(umask 077; printf '{"lane":"%s","vendor":"%s","session_mode":"%s","idle_timeout":%s,"model":"%s","effort":"%s","timeout":%s,"job_timeout":%s,"mode":"%s","workdir":"%s","foreman_session":"%s","candidate":"%s/%s","started":"%s"}\n' \
+  "$(json_escape "$LANE")" "$(json_escape "$VENDOR")" "$(json_escape "$SESSION_MODE")" \
+  "$IDLE_TIMEOUT" "$(json_escape "$MODEL")" "$(json_escape "$EFFORT")" \
+  "$TIMEOUT" "$JOB_TIMEOUT_JSON" "$(json_escape "$MODE")" \
   "$(json_escape "$WORKDIR")" "$(json_escape "$FOREMAN_SESSION")" \
   "$RESOLVED_IDX" "$RESOLVED_TOTAL" \
   "$(date -u +%FT%TZ)" > "$JOB_DIR/meta.json")

package/scripts/doctor.sh

@@ -31,6 +31,7 @@ done
 REPO="${OMNILANE_DOCTOR_REPO:-$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)}"
 OMNILANE_HOME="${OMNILANE_HOME:-$HOME/.omnilane}"
 PROBE_SCRIPT="${OMNILANE_PROVIDER_PROBE_SCRIPT:-$REPO/scripts/provider-probe.sh}"
+GOAL_LOOP="${OMNILANE_DOCTOR_GOAL_LOOP:-$REPO/scripts/lib/goal-loop.sh}"
 PASS_COUNT=0
 WARN_COUNT=0
 FAIL_COUNT=0
@@ -123,6 +124,31 @@ else
   report PASS state "$OMNILANE_HOME is accessible"
 fi
 
+goals_store="$OMNILANE_HOME/goals"
+if [[ ! -f "$GOAL_LOOP" || -L "$GOAL_LOOP" ]]; then
+  report FAIL goal-orchestrator "$GOAL_LOOP must be a regular file"
+elif ! /bin/bash -n "$GOAL_LOOP" 2>/dev/null; then
+  report FAIL goal-orchestrator "$GOAL_LOOP has invalid bash syntax"
+elif [[ -L "$goals_store" || ( -e "$goals_store" && ! -d "$goals_store" ) ]]; then
+  report FAIL goal-orchestrator "$goals_store must be a real directory, not a symlink or file"
+elif [[ -d "$goals_store" ]]; then
+  goals_mode="$(stat -f '%Lp' "$goals_store" 2>/dev/null || true)"
+  if [[ ! "$goals_mode" =~ ^[0-7]{3,4}$ ]]; then
+    goals_mode="$(stat -c '%a' "$goals_store" 2>/dev/null || true)"
+  fi
+  if [[ "$goals_mode" =~ ^0?700$ ]]; then
+    report PASS goal-orchestrator "goal-loop.sh parses; goals store mode is $goals_mode"
+  elif [[ -n "$goals_mode" ]]; then
+    report WARN goal-orchestrator \
+      "goal-loop.sh parses; goals store mode is $goals_mode, owner-only 700 is safer"
+  else
+    report WARN goal-orchestrator \
+      "goal-loop.sh parses; could not determine goals store permissions"
+  fi
+else
+  report PASS goal-orchestrator "goal-loop.sh parses; goals store is not present yet"
+fi
+
 completion_plugin_state() {
   local config_dir state
   config_dir="${CLAUDE_CONFIG_DIR:-${HOME:-}/.claude}"

package/scripts/jobs.sh

@@ -10,6 +10,8 @@ set -euo pipefail
 # Runtime-relative shared library.
 # shellcheck disable=SC1091
 source "$(dirname "${BASH_SOURCE[0]}")/lib/common.sh"
+# shellcheck disable=SC1091
+source "$OMNILANE_REPO/scripts/lib/live-protocol.sh"
 JOBS="$OMNILANE_HOME/jobs"
 JOB_ID_PATTERN='^[0-9]{8}-[0-9]{6}-[0-9]+-[0-9]+$'
 JSON_MODE=0
@@ -145,15 +147,28 @@ read_job_pid() {
 }
 
 load_job_vendor() {
+  local session_mode_re='"session_mode":"(live|single-shot)"'
   if ! read_public_metadata "$JOB_DIR/meta.json" || ! parse_stats_metadata "$PUBLIC_METADATA"; then
     die 1 "job metadata is not safely readable"
   fi
   JOB_VENDOR="$STATS_VENDOR"
+  JOB_SESSION_MODE=""
+  if [[ "$PUBLIC_METADATA" =~ $session_mode_re ]]; then
+    JOB_SESSION_MODE="${BASH_REMATCH[1]}"
+  elif [[ "$JOB_VENDOR" == "claude" ]]; then
+    # Compatibility with live jobs created before session_mode metadata.
+    JOB_SESSION_MODE="live"
+  else
+    JOB_SESSION_MODE="single-shot"
+  fi
 }
 
 require_live_job() {
   load_job_vendor
-  [[ "$JOB_VENDOR" == "claude" ]] || die 1 "job vendor '$JOB_VENDOR' is not live-capable; it runs in single-shot mode"
+  live_vendor_capable "$JOB_VENDOR" ||
+    die 1 "job vendor '$JOB_VENDOR' is not live-capable (live vendors: $(live_capable_vendors)); it runs in single-shot mode"
+  [[ "$JOB_SESSION_MODE" == "live" ]] ||
+    die 1 "job vendor '$JOB_VENDOR' was dispatched in single-shot mode"
 }
 
 wait_for_live_ready() {
@@ -204,10 +219,11 @@ parse_stats_metadata() {
 }
 
 parse_audit_metadata() {
-  local value="$1" metadata_re
+  local value="$1" current_re legacy_re
   local json_string='([^"\\]|\\["\\/bfnrt]|\\u[0-9a-fA-F]{4})*'
-  metadata_re="^\\{\"lane\":\"[a-z][a-z0-9-]*\",\"vendor\":\"(${OMNILANE_VENDOR_ALT})\",\"model\":\"${json_string}\",\"effort\":\"${json_string}\",\"timeout\":(0|[1-9][0-9]*),\"job_timeout\":(null|0|[1-9][0-9]*),\"mode\":\"(advise|work)\",\"workdir\":\"${json_string}\",\"candidate\":\"[1-9][0-9]*/[1-9][0-9]*\",\"started\":\"[0-9]{4}-[0-9]{2}-[0-9]{2}T[0-9]{2}:[0-9]{2}:[0-9]{2}Z\"\\}$"
-  [[ "$value" =~ $metadata_re ]]
+  current_re="^\\{\"lane\":\"[a-z][a-z0-9-]*\",\"vendor\":\"(${OMNILANE_VENDOR_ALT})\",\"session_mode\":\"(live|single-shot)\",\"idle_timeout\":(0|[1-9][0-9]*),\"model\":\"${json_string}\",\"effort\":\"${json_string}\",\"timeout\":(0|[1-9][0-9]*),\"job_timeout\":(null|0|[1-9][0-9]*),\"mode\":\"(advise|work|sysops)\",\"workdir\":\"${json_string}\",\"foreman_session\":\"${json_string}\",\"candidate\":\"[1-9][0-9]*/[1-9][0-9]*\",\"started\":\"[0-9]{4}-[0-9]{2}-[0-9]{2}T[0-9]{2}:[0-9]{2}:[0-9]{2}Z\"\\}$"
+  legacy_re="^\\{\"lane\":\"[a-z][a-z0-9-]*\",\"vendor\":\"(${OMNILANE_VENDOR_ALT})\",\"model\":\"${json_string}\",\"effort\":\"${json_string}\",\"timeout\":(0|[1-9][0-9]*),\"job_timeout\":(null|0|[1-9][0-9]*),\"mode\":\"(advise|work|sysops)\",\"workdir\":\"${json_string}\",(\"foreman_session\":\"${json_string}\",)?\"candidate\":\"[1-9][0-9]*/[1-9][0-9]*\",\"started\":\"[0-9]{4}-[0-9]{2}-[0-9]{2}T[0-9]{2}:[0-9]{2}:[0-9]{2}Z\"\\}$"
+  [[ "$value" =~ $current_re || "$value" =~ $legacy_re ]]
 }
 
 print_stat_counts() {
@@ -352,8 +368,7 @@ case "${1:-}" in
     else
       foreman_value="${foreman_session-}"
     fi
-    payload="$(printf '{"type":"user","omnilane_job_id":"%s","foreman_session":"%s","message":{"role":"user","content":[{"type":"text","text":"%s"}]}}' \
-      "$(json_escape "$2")" "$(json_escape "$foreman_value")" "$(json_escape "$3")")"
+    payload="$(live_encode_message "$JOB_VENDOR" "$2" "$foreman_value" "$3")"
     set +e
     write_fifo_with_timeout "$JOB_DIR/inbox.fifo" "$payload"
     send_rc=$?
@@ -404,11 +419,7 @@ case "${1:-}" in
     done
     read_exit_code "$JOB_DIR/exit" || die 1 "invalid recorded exit status"
     events_path="$JOB_DIR/events.jsonl"
-    result_count=0
-    if [[ -f "$events_path" && ! -L "$events_path" ]]; then
-      result_count="$(grep -Ec '"type"[[:space:]]*:[[:space:]]*"result"' "$events_path" || true)"
-      [[ "$result_count" =~ ^[0-9]+$ ]] || result_count=0
-    fi
+    result_count="$(live_count_result_events "$JOB_VENDOR" "$events_path")"
     echo "closed $2 (result_events=$result_count exit=$RECORDED_EXIT)"
     exit "$RECORDED_EXIT"
     ;;
@@ -565,7 +576,8 @@ case "${1:-}" in
       elif [[ -d "$artifact" ]]; then
         audit_emit "$id" nested-directory
         findings=$((findings + 1)); job_failed=1
-      elif [[ -p "$artifact" && "${artifact##*/}" == "inbox.fifo" ]]; then
+        elif [[ -p "$artifact" && ( "${artifact##*/}" == "inbox.fifo" ||
+                                    "${artifact##*/}" == "runner-inbox.fifo" ) ]]; then
         mode="$(path_mode "$artifact" 2>/dev/null || true)"
         if [[ "$mode" != "600" ]]; then
           audit_emit "$id" unsafe-fifo-mode