codex-councillisted

Multi-perspective Codex review via parallel `codex exec` sub-agents. Claude (the orchestrator) composes the role panel on-the-fly per invocation from the user's actual work; there is no built-in role catalog and no domain list. Auto-use ONLY when the user's text contains the phrase "codex council" or "codex team reconciliation" (and close variants such as "ask codex council," "codex council review," "reconcile with codex team"). Otherwise stop — broader phrases like "codex agent team," "codex panel," "fan out to codex agents," "agent team," "agents in parallel," "subagents," "council review," "panel review," "multi-angle review" route to Claude Code's built-in `Agent` tool. Direct invocation `/codex-council:codex-council`. On invocation, ultrathink about what the user is actually doing, compose a tailored panel ground-up, announce it briefly, then launch without a manual approval gate.
ehzawad/codex-council · ★ 1 · DevOps & Infrastructure · score 55

Install: claude install-skill ehzawad/codex-council

# Codex Council Fan out a prompt to N parallel `codex exec` sub-agents, each framed with a role tailored to the work. The script aggregates all responses into one structured markdown report. Each role keeps its own Codex thread per project so framings accumulate across calls. Codex is strongest on technical and structured reasoning. Whether a council adds value over a single pass should be judged from the current task, not from a label on it. **You (Claude) are the orchestrator.** The panel-proposal step is load-bearing: **ultrathink** there. Read the user's actual work, figure out what judgment they need, design role ids / labels / instructions ground-up from that specific work. There is no catalog, no checklist of domains, no template panel to reach for. Treat your composed task-specific panel as granted by default: announce it briefly, then trigger `codex exec` without asking for launch approval. ## Disambiguation gate — only if "codex council" / "codex team" is missing If the trigger phrase that fired this skill does **not** contain "codex council" or "codex team reconciliation" (or a close variant), **stop**. The user may have meant Claude Code's built-in `Agent` tool (subagents spawned with `subagent_type` like `general-purpose`, `Plan`, `Explore`, `claude-code-guide`, `code-reviewer`) — a different mechanism (Claude's model with direct tool access, not OpenAI Codex via `codex exec`). Ask one short disambiguation via `AskUserQuestion` before doing anything: - Que