{"slug": "council-five-independent-advisors-for-your-agent", "title": "Council – Five independent advisors for your agent", "summary": "Developer Matt Akins released Council, an open-source skill that runs five independent AI advisors in parallel and has a Chairman subagent synthesize their verdicts into one recommendation. Council works with Claude Code, OpenAI Codex, OpenCode, Cursor, and ChatGPT Work, installs via plugin marketplaces or the command `npx skills add mattakins/council -g`, and supports 3- or 7-advisor runs, a research mode, and a `quick` flag that skips the Chairman. The project is inspired by Andrej Karpathy's LLM Council and targets architectural, product, and naming decisions.", "body_md": "**Five independent advisors. One clear recommendation.**\n\nCouncil is a skill that helps you make decisions. Ask a question about your project. Five independent advisors weigh in. The Chairman brings it all together. Your agent adds its own take and next step.\n\nUse it for architectural calls, product and marketing decisions, naming debates — any decision worth more than one perspective.\n\n[Website](https://mattakins.github.io/Council/) · [Install](#install) · [See examples](https://mattakins.github.io/Council/#examples)\n\nWorks with Codex, Claude Code, OpenCode, Cursor, ChatGPT Work, and other harnesses that support parallel, isolated subagents.\n\nInspired by [Karpathy's LLM Council](https://github.com/karpathy/llm-council).\n\n```\nClaude: /council:council should we rewrite this service in Rust?\nCodex:  $council:council deep what's the commercial viability of this app right now?\n/plugin marketplace add mattakins/council\n/plugin install council@mattakins\n```\n\nPlugin skills are namespaced. Invoke this skill as `/council:council <question>`.\n\n```\ncodex plugin marketplace add mattakins/council\ncodex plugin add council@mattakins\n```\n\nInvoke the plugin skill as `$council:council <question>`.\n\n```\nnpx skills add mattakins/council -g\n```\n\nThe Skills CLI installs the shared `SKILL.md` into the paths expected by the agents you select.\n\nClone or copy `plugins/council/skills/council` into one of these supported locations:\n\n- Claude Code: `~/.claude/skills/council/`\n- Codex: `~/.agents/skills/council/`\n- OpenCode: `~/.agents/skills/council/` or`~/.config/opencode/skills/council/`\n\n| Command | What it does | \n|---|---|\n| `<council> <question>` | Default: 5 agents + Chairman | \n| `<council> 3 <question>` | Use 3 agents | \n| `<council> 7 <question>` | Use 7 agents | \n| `<council> deep <question>` | Research + 7 agents + Chairman (full treatment) | \n| `<council> research <question>` | Explicitly enable research | \n| `<council> no research <question>` | Skip research, reasoning only | \n| `<council> \"SEO, legal, UX\" <question>` | Specify perspectives explicitly | \n| `<council> quick <question>` | Skip Chairman, faster output | \n| `<council> no chairman <question>` | Same as quick | \n| `<council> save <question>` | Write output to `council-YYYY-MM-DD-<slug>.md` | \n| `<council> opus <question>` | Run all advisors on Opus (any named model works if your harness supports per-subagent overrides) | \n| `<council> sonnet <question>` | Run all advisors on Sonnet | \n\nReplace `<council>` with `/council:council` for the Claude plugin or `$council:council` for the Codex plugin. Flags combine, and options also work as plain language — e.g. `use 4 advisors, no research, and save: <question>`.\n\n1. **Parse** — extracts agent count, perspectives, flags, and auto-detects if research is needed\n2. **Research detection** — if the question involves markets, companies, current events, or real-world data, agents are automatically told to research before answering. No flag needed.\n3. **Parallel agents** — all advisors run simultaneously, blind to each other, each from their assigned perspective\n4. **Chairman synthesis** — a fresh, isolated subagent synthesizes the council into a definitive verdict (skippable with`quick` )\n5. **My Take** — the final section: an opinionated recommendation drawing on both the council outputs and full conversation context\n\n```\nChairman's Verdict:\n[Definitive synthesis]\n\n| Agent | Perspective | Recommendation | Confidence |\n| ...   | ...         | ...            | ...        |\n\nConsensus: ...\nDissent (if any): ...\nRisks: ...\nVerdict: ...\n\nMy Take: ...        ← always present, always opinionated\nNext step: ...\n```\n\n| Tool | Works? | Install path | \n|---|---|---|\n| Claude Code | ✅ | Plugin or `~/.claude/skills/council/` | \n| OpenAI Codex | ✅ | Plugin or `~/.agents/skills/council/` | \n| OpenCode | ✅ | `~/.agents/skills/council/` or native OpenCode path | \n| Cursor | ✅ | Agent Skills install ( `npx skills add` ) | \n| Other harnesses | Conditional | Requires parallel subagent support | \n\nThe instruction format follows the open Agent Skills specification. Full behavior requires a harness that can launch parallel, isolated subagents.\n\nMIT", "url": "https://wpnews.pro/news/council-five-independent-advisors-for-your-agent", "canonical_source": "https://github.com/mattakins/Council", "published_at": "2026-09-13 15:00:52+00:00", "updated_at": "2026-09-13 15:15:18.534464+00:00", "lang": "en", "topics": ["ai-agents", "ai-tools", "ai-products", "developer-tools"], "entities": ["Council", "Matt Akins", "Claude Code", "OpenAI Codex", "OpenCode", "Cursor", "ChatGPT Work", "Andrej Karpathy"], "alternates": {"html": "https://wpnews.pro/news/council-five-independent-advisors-for-your-agent", "markdown": "https://wpnews.pro/news/council-five-independent-advisors-for-your-agent.md", "text": "https://wpnews.pro/news/council-five-independent-advisors-for-your-agent.txt", "jsonld": "https://wpnews.pro/news/council-five-independent-advisors-for-your-agent.jsonld"}}