Skip to content
Back to search

council-of-high-intelligence

Structured multi-perspective deliberation for hard decisions. Run full councils, focused triads, or duo debates across Claude Code, Codex, Gemini CLI, and OpenCode.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked yesterday

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/0xNyk/council-of-high-intelligence with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/0xNyk/council-of-high-intelligence'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/0xnyk--council-of-high-intelligence/badge.svg)](/listings/0xnyk--council-of-high-intelligence)

HTML

<a href="/listings/0xnyk--council-of-high-intelligence"><img src="/api/listings/0xnyk--council-of-high-intelligence/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

./install.sh                         # Claude Code
./install.sh --codex                 # Claude Code + Codex
./install.sh --codex-only            # Codex only
./install.sh --gemini                # Claude Code + Gemini CLI
./install.sh --gemini-only           # Gemini CLI only
./install.sh --opencode              # Claude Code + OpenCode
./install.sh --opencode-only         # OpenCode only
./install.sh --copy-configs          # Include provider-routing templates
./install.sh --dry-run               # Preview writes

Custom target directories are supported through `--claude-dir`, `--codex-dir`, `--gemini-dir`, and `--opencode-dir`. Run `./install.sh --help` for the current contract.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 16 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. {
  2. "name": "council-of-high-intelligence",
  3. "description": "Official marketplace for the Council of High Intelligence — multi-persona, multi-provider deliberation for Claude Code.",
  4. "owner": {
  5. "name": "0xNyk",
  6. "url": "https://github.com/0xNyk"
  7. },
  8. "plugins": [
  9. {
  10. "name": "council",
  11. "source": "./",
  12. "description": "18 AI personas deliberate your hardest decisions — structured multi-round deliberation with genuine model diversity. One command: /council",
  13. "version": "1.2.0"
  14. }
  15. ]
  16. }
README.md · README · 260 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. # Council of High Intelligence
  2. <p align="center">
  3. <img src="assets/header.jpeg" alt="Council of High Intelligence" width="800">
  4. </p>
  5. <p align="center">
  6. Structured multi-perspective deliberation for decisions that deserve more than one reasoning path.
  7. </p>
  8. <p align="center">
  9. <a href="https://github.com/0xNyk/council-of-high-intelligence/actions/workflows/lint.yml"><img src="https://github.com/0xNyk/council-of-high-intelligence/actions/workflows/lint.yml/badge.svg" alt="Lint"></a>
  10. <a href="https://github.com/0xNyk/council-of-high-intelligence/releases"><img src="https://img.shields.io/github/v/release/0xNyk/council-of-high-intelligence" alt="Release"></a>
  11. <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-blue" alt="License"></a>
  12. </p>
  13. <p align="center">
  14. Claude Code · Codex · Gemini CLI · OpenCode
  15. </p>
  16. The council assigns a hard question to deliberately different analytical personas, keeps
  17. their first positions independent, forces direct disagreement, and returns a verdict that
  18. preserves unresolved questions, dissent, kill criteria, and the next concrete action.
  19. ## Try a council
  20. Claude Code users can install the plugin directly:
  21. ```text
  22. /plugin marketplace add 0xNyk/council-of-high-intelligence
  23. /plugin install council@council-of-high-intelligence
  24. ```
  25. Then convene the full council, a faster panel, or a two-member dialectic:
  26. ```text
  27. /council Should we open-source our agent framework?
  28. /council --quick Should we add caching here?
  29. /council --duo Should we use microservices or a monolith?
  30. ```
  31. For Codex, Gemini CLI, or OpenCode, clone once and use the matching installer flag:
  32. ```bash
  33. git clone https://github.com/0xNyk/council-of-high-intelligence.git
  34. cd council-of-high-intelligence
  35. ./install.sh --codex-only
  36. ./install.sh --gemini-only
  37. ./install.sh --opencode-only
  38. ```
  39. Restart the target client after installation. The command remains `/council` on every
  40. supported host.
  41. ## Use it for the decision boundary
  42. Good council questions have material downside, competing values, incomplete evidence, or
  43. an irreversible choice. Write down the decision, constraints, evidence, reversibility, and
  44. deadline before adding personas.
  45. ![Decision field notes](assets/decision-field-notes.jpeg)
  46. The field notes distinguish facts from inference, assumptions, and unknowns. They also make
  47. it harder to use a long deliberation as decoration for a decision already made.
  48. ### When to use something smaller
  49. - Use a direct answer or primary documentation for factual lookups.
  50. - Run an experiment when the choice is cheap and reversible.
  51. - Use `--quick` when cross-examination will not change the outcome.
  52. - Use `--duo` when one tension matters more than broad coverage.
  53. - Do not convene a council to manufacture support for a preferred answer.
  54. ## Choose a mode
  55. | Mode | Shape | Use it when |
  56. |---|---|---|
  57. | Full | Independent analysis, cross-examination, final stance, synthesis | Stakes are high and competing frames need contact |
  58. | Quick | Restate, rapid analysis, final positions | The decision needs breadth but not a full adversarial round |
  59. | Duo | Opening positions, direct response, final statements | One polarity defines the decision |
  60. ![Council mode selector](assets/mode-selector.jpeg)
  61. ```text
  62. /council --full What is the right pricing model?
  63. /council --quick --triad shipping Should we release today?
  64. /council --duo --members torvalds,ada Is this abstraction worth it?
  65. ```
  66. Named triads select three relevant lenses without assembling the whole council:
  67. ```text
  68. /council --triad strategy Where is our defensible advantage?
  69. /council --triad risk What could make this launch irreversible?
  70. /council --triad ai-product Which capability belongs in the product?
  71. ```
  72. Available domains include `architecture`, `strategy`, `ethics`, `debugging`, `risk`,
  73. `shipping`, `product`, `founder`, `ai`, `ai-product`, `ai-safety`, `decision`, `systems`,
  74. `uncertainty`, `design`, `economics`, and `bias`. The canonical routing table lives in
  75. [`SKILL.md`](SKILL.md).
  76. ## What the protocol protects
  77. The full protocol has a fixed round budget. Members first restate the problem, analyze
  78. blind, cross-examine other positions, declare a final stance, and then enter synthesis.
  79. Enforcement checks look for premature agreement, repeated claims, missing dissent, and
  80. unsupported confidence.
  81. ![Verdict blueprint](assets/verdict-blueprint.jpeg)
  82. Verdicts lead with what remains unresolved. A recommendation is paired with acceptable
  83. compromises, kill criteria, and one concrete next step. When the weighted tally remains
  84. split, the result returns that split to the user instead of turning it into prose consensus.
  85. Evidence labels separate:
  86. - `FACT`: directly supported by supplied or retrieved evidence;
  87. - `INFERENCE`: follows from evidence but is not directly observed;
  88. - `ASSUMPTION`: required for the argument and still unverified;
  89. - `UNKNOWN`: missing information that could change the decision.
  90. ## Track the outcome
  91. A verdict is useful only if it can be checked later. Before acting, record the prediction,
  92. owner, review date, and evidence that would change the recommendation. At the checkpoint,
  93. mark the result confirmed, revised, reversed, or inconclusive instead of rewriting the
  94. original rationale.
  95. ![Council outcome ledger](assets/outcome-ledger.jpeg)
  96. ## The 18 lenses
  97. Members are analytical instruments, not impersonation claims. Each persona has a grounding
  98. protocol, a method, known blind spots, and a structured response contract.
  99. ![Council panel composition](assets/panel-composition.jpeg)
  100. | Member | Primary lens | Useful counterweight |
  101. |---|---|---|
  102. | Aristotle | Categories and structure | Lao Tzu on emergence and excess structure |
  103. | Socrates | Assumption destruction | Feynman on reconstruction from first principles |
  104. | Sun Tzu | Terrain and adversarial strategy | Aurelius on internal control and moral cost |
  105. | Ada Lovelace | Formal systems and abstraction | Machiavelli on incentives and informal power |
  106. | Marcus Aurelius | Resilience and moral clarity | Sun Tzu on external competition |
  107. | Machiavelli | Power and incentives | Ada on formal consistency |
  108. | Lao Tzu | Non-action and emergence | Aristotle on explicit categories |
  109. | Richard Feynman | Explanation and empirical debugging | Socrates on the premise itself |
  110. | Linus Torvalds | Shipping and maintainability | Meadows on system-level consequences |
  111. | Miyamoto Musashi | Timing and decisive action | Torvalds on acting before the ideal moment |
  112. | Alan Watts | Reframing and false problems | Torvalds on concrete implementation |
  113. | Andrej Karpathy | Empirical ML behavior | Sutskever on frontier risk |
  114. | Ilya Sutskever | Scaling and AI safety | Karpathy on observation and iteration |
  115. | Daniel Kahneman | Cognitive bias | Feynman on explicit causal reasoning |
  116. | Donella Meadows | Feedback loops and high-impact interventions | Torvalds on local fixes |
  117. | Charlie Munger | Inversion and model lattices | Aristotle on single-system classification |
  118. | Nassim Taleb | Tail risk and fragility | Karpathy on smooth empirical trends |
  119. | Dieter Rams | User clarity and restraint | Ada on what can be formalized |
  120. The repository stores each member contract under [`agents/`](agents/). Custom panels can
  121. use `--members`, named triads, or the `classic`, `exploration-orthogonal`, and
  122. `execution-lean` profiles documented in [`SKILL.md`](SKILL.md).
  123. ## Multi-provider routing
  124. The detection script checks which supported providers are available, then the coordinator
  125. distributes seats across them. Polarity pairs are separated when possible so a single model
  126. family does not play both sides of a disagreement.
  127. ![Council provider routing](assets/provider-routing.jpeg)
  128. | Provider path | Detection |
  129. |---|---|
  130. | Native host subagents | Available through the active supported client |
  131. | OpenAI | `codex` executable |
  132. | Google | `gemini` executable |
  133. | Ollama | `ollama` executable |
  134. | NVIDIA NIM | `NVIDIA_API_KEY` |
  135. | Cursor | `cursor-agent` executable or configured login |
  136. Preview routing without running a council:
  137. ```text
  138. /council --dry-route --triad decision Should we accept this acquisition offer?
  139. ```
  140. Use `--no-auto-route` to keep native-host defaults. Use `--models <path>` with a copy of
  141. [`configs/provider-model-slots.example.yaml`](configs/provider-model-slots.example.yaml)
  142. for an explicit seat map. Provider failure is reported before the seat falls back to the
  143. native host.
  144. ## Installation reference
  145. ```bash
  146. ./install.sh # Claude Code
  147. ./install.sh --codex # Claude Code + Codex
  148. ./install.sh --codex-only # Codex only
  149. ./install.sh --gemini # Claude Code + Gemini CLI
  150. ./install.sh --gemini-only # Gemini CLI only
  151. ./install.sh --opencode # Claude Code + OpenCode
  152. ./install.sh --opencode-only # OpenCode only
  153. ./install.sh --copy-configs # Include provider-routing templates
  154. ./install.sh --dry-run # Preview writes
  155. ```
  156. Custom target directories are supported through `--claude-dir`, `--codex-dir`,
  157. `--gemini-dir`, and `--opencode-dir`. Run `./install.sh --help` for the current contract.
  158. ## Verify the checkout
  159. ```bash
  160. ./scripts/council-simulation-checklist.sh
  161. ./install.sh --dry-run --codex
  162. ./install.sh --dry-run --gemini
  163. ./install.sh --dry-run --opencode
  164. ```
  165. The checklist validates persona structure, host-protocol parity, routing configuration,
  166. execution checkpoints, verdict fields, and installer behavior. Use the
  167. [demo session pack](demos/session-pack.md) to exercise full, quick, and duo modes.
  168. ## Repository map
  169. ```text
  170. SKILL.md canonical coordinator protocol
  171. SKILL.codex.md Codex host mirror
  172. SKILL.gemini.md Gemini CLI host mirror
  173. SKILL.opencode.md OpenCode host mirror
  174. agents/ 18 grounded persona contracts
  175. configs/ provider and model-routing examples
  176. demos/ sample sessions and verdict template
  177. scripts/ detection, conversion, and validation tools
  178. assets/ brand system and README visuals
  179. install.sh multi-host installer
  180. ```
  181. Protocol changes start in `SKILL.md` and must remain behaviorally aligned with each host
  182. mirror. See [CONTRIBUTING.md](CONTRIBUTING.md) for the required checks and review contract.
  183. Security reports belong in [SECURITY.md](SECURITY.md), not a public issue.
  184. ## Project links
  185. - [Release history](CHANGELOG.md)
  186. - [Brand kit](assets/BRAND.md)
  187. - [OSS profile](https://www.nyk.dev/oss/council-of-high-intelligence)
  188. - [Sponsor ongoing maintenance](https://github.com/sponsors/0xNyk)
  189. - [Maintainer: Nyk](https://nyk.dev) · [@nykdotdev](https://x.com/nykdotdev)
  190. - [Issues](https://github.com/0xNyk/council-of-high-intelligence/issues)
  191. <p align="center">
  192. <picture>
  193. <source media="(prefers-color-scheme: dark)" srcset="assets/star-history-dark.svg">
  194. <img src="assets/star-history-light.svg" alt="Council of High Intelligence star history" width="600">
  195. </picture>
  196. </p>
  197. ## License
  198. [MIT](LICENSE) © 2026 [0xNyk](https://github.com/0xNyk)

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.