Skip to content

scdenney/open-science-skills / codex/orchestrate

orchestrate

Verified

Orchestrate complex work from an active gpt-6-astra or gpt-6.1-sol Codex session. The detected lead owns decomposition, integration, and verification; Astra keeps the hardest reasoning in the lead, while Sol escalates unusually difficult units to Astra. Both route bounded work to Sol at medium effort, mechanical work to Luna, and can use a cross-vendor Claude peer. Not for routine single-context work. Use when the user explicitly asks to orchestrate, delegate, fan out, parallelize, assign subagents, obtain independent checks, or have Codex act as tech lead.

FileKindSize
SKILL.mdskill md21.2 KB
agents/openai.yamlconfig251 B
assets/architecture.svgother4.8 KB
scripts/check-lead-runtime.shexecutablescript3.1 KB
scripts/claude-peer.shexecutablescript2.7 KB

scripts/claude-peer.sh

View on GitHub
#!/usr/bin/env bash
# claude-peer.sh — invoke Claude Code (Anthropic) as a cross-vendor peer for /orchestrate.
#
# Runs a separate Claude CLI consult using a self-contained prompt.
# The caller may pair it with an OpenAI worker for a blind cross-vendor review.
# Different vendors can still share errors; compare their checkable evidence.
# Parent sandbox, authentication, and network permissions apply.
#
# Usage:
#   claude-peer.sh [-C DIR] [--timeout SEC] [--model ID] [--effort LEVEL]
#                  [--out FILE] (--prompt TEXT | --prompt-file PATH | -)
#
#   -C DIR          working dir Claude sees (default: $PWD)
#   --timeout SEC   hard kill after SEC seconds (default: 600)
#   --model ID      default: fable; select a different Claude model explicitly
#   --effort LEVEL  reasoning effort, passed to claude as --effort LEVEL
#                   (default: high)
#   --out FILE      also tee Claude's stdout+stderr here (for background reads)
#   --prompt TEXT   prompt as a single argument
#   --prompt-file P read prompt from file P
#   -               read prompt from stdin
set -euo pipefail

DIR="$PWD"; TIMEOUT=600; MODEL="fable"; EFFORT="high"; OUT=""; PROMPT=""; PROMPT_SET=0

die(){ echo "claude-peer: $*" >&2; exit 2; }

while [ $# -gt 0 ]; do
  case "$1" in
    -C|--dir)      DIR="${2:?}"; shift 2 ;;
    --timeout)     TIMEOUT="${2:?}"; shift 2 ;;
    --model)       MODEL="${2:?}"; shift 2 ;;
    --effort)      EFFORT="${2:?}"; shift 2 ;;
    --out)         OUT="${2:?}"; shift 2 ;;
    --prompt)      PROMPT="${2:?}"; PROMPT_SET=1; shift 2 ;;
    --prompt-file) PROMPT="$(cat "${2:?}")"; PROMPT_SET=1; shift 2 ;;
    -)             PROMPT="$(cat)"; PROMPT_SET=1; shift ;;
    -h|--help)     grep '^#' "$0" | sed 's/^# \{0,1\}//'; exit 0 ;;
    *)             die "unknown argument: $1" ;;
  esac
done

[ "$PROMPT_SET" -eq 1 ] || die "no prompt — pass --prompt, --prompt-file, or -"
[ -n "$PROMPT" ] || die "empty prompt rejected"

command -v claude >/dev/null 2>&1 || die "claude CLI not found on PATH"

run() {
  cd "$DIR"
  local cmd=(claude -p "$PROMPT" --model "$MODEL" --output-format text)
  if [ -n "$EFFORT" ]; then cmd+=(--effort "$EFFORT"); fi
  # `timeout` is not preinstalled on macOS (only via GNU coreutils) — guard
  # rather than assume, matching codex-peer.sh's own fix for the identical
  # problem (confirmed missing on this machine 2026-07-18: neither `timeout`
  # nor `gtimeout` is on PATH).
  if command -v timeout >/dev/null 2>&1; then
    timeout "${TIMEOUT}s" "${cmd[@]}" < /dev/null
  elif command -v gtimeout >/dev/null 2>&1; then
    gtimeout "${TIMEOUT}s" "${cmd[@]}" < /dev/null
  else
    "${cmd[@]}" < /dev/null
  fi
}

if [ -n "$OUT" ]; then
  run 2>&1 | tee "$OUT"
else
  run
fi
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall