Council

Name: Council
Author: boshu2

boshu2/agentops

2.9k installs
416 repo stars
Updated August 5, 2026
boshu2/agentops

council is an agentops skill that runs independent multi-judge consensus before committing to high-stakes or irreversible architecture decisions.

About

Council is an agentops judgment skill that routes high-stakes decisions through independent multi-judge consensus instead of a single model answer. The fleet stub documents absorbed triggers from multi-model triangulation for cross-validating competing approaches with Codex, Gemini, and Grok judges, and from cross-vendor-trust-gate for operating trust-gate.sh and enforcing cross-vendor parity via skill.trust.json. Mixed-model panels can use dual-pane-atm durable Opus plus Codex panes, while reverse-engineer routes one-way-door steal decisions into council. Outputs bind to result.json and verdict.json schemas with a shared standards dependency and isolated context window excluding history. The canonical implementation moved to Mount Olympus in 2026-06-10; this registry stub preserves fleet routing until the agentops catalog closer updates skill-prune Lane A. Runtime contract references claude-code-latest-features.md for multi-judge runs on the shared Claude surface. Examples include /council prompts for policy engine swaps like Cedar migration decisions.

Multi-judge consensus for irreversible architecture forks and one-way-door decisions.
Absorbs multi-model triangulation and cross-vendor-trust-gate trigger surfaces.
Outputs result.json and verdict.json under a shared standards dependency contract.
Mixed-model panels route through dual-pane-atm Opus plus Codex durable panes.
Canonical body lives at Mount Olympus; this stub preserves fleet routing until catalog prune.

Council by the numbers

2,858 all-time installs (skills.sh)
+33 installs in the week ending Aug 5, 2026 (Skillselion tracking)
Ranked #276 of 16,546 AI & Agent Building skills by installs in the Skillselion catalog
Security screen: MEDIUM risk (skills.sh audit)
Data as of Aug 5, 2026 (Skillselion catalog sync)

At a glance

council capabilities & compatibility

Capabilities: multi judge independent verdict machinery for hi · cross model triangulation across codex, gemini, · cross vendor trust gate enforcement via trust ga · mixed model panel routing through dual pane atm
Use cases: orchestration · planning · security audit

From the docs

What council says it does

Run multi-judge consensus. Use when: an irreversible or high-stakes decision needs independent judges

SKILL.md

npx skills add https://github.com/boshu2/agentops --skill council

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

[![Listed on Skillselion](https://skillselion.com/badge/skills/boshu2/agentops/council.svg)](https://skillselion.com/skills/boshu2/agentops/council)

Installs	2.9k
repo stars	★ 416
Security audit	3 / 3 scanners passed
Last updated	August 5, 2026
Repository	boshu2/agentops ↗

How do I get independent judge verdicts before committing to an irreversible architecture fork or one-way policy decision?

Run multi-judge consensus with independent verdicts before irreversible architecture forks, one-way doors, or high-stakes scoring decisions.

Who is it for?

Teams running agentops workflows who need cross-model triangulation on architecture forks and trust-gated verdicts.

Skip if: Skip when you need the full implementation scripts; the canonical body moved to Mount Olympus and this entry is a routing stub.

When should I use this skill?

User asks for /council, multi-judge consensus, cross-vendor trust gate, or a second opinion on high-stakes design choices.

What you get

Structured verdict.json and result.json outputs from a multi-judge panel with cross-vendor trust gating when required.

result.json
verdict.json

By the numbers

Produces 2 JSON artifacts: result.json and verdict.json
Declares skill_api_version 1 with standards skill dependency
Absorbs 2 folded trigger surfaces: multi-model-triangulation and cross-vendor-trust-gate

Files

SKILL.mdMarkdownGitHub ↗

council — moved to Mount Olympus (2026-06-10)

This skill encodes independent-verdict machinery and now lives with the outer gate product. Canonical: ~/dev/mt-olympus/.claude/skills/council/SKILL.md — read and follow that file. This stub preserves fleet routing until the using-agentops catalog closer updates the registry (skill-prune Lane A, evidence/skill-prune-recon.md).

Absorbed trigger surfaces (skill-prune phase 2)

Council also fires for the use-cases of two folded-in skills:

multi-model-triangulation — cross-validate decisions using multiple AI

models (Codex, Gemini, Grok). Use when asked to "get a second opinion", when evaluating competing approaches, or for high-stakes decisions: run the question through council's independent judges instead of a single model.

cross-vendor-trust-gate — run the skill-factory final trust gate:

operate trust-gate.sh, read skill.trust.json (trust_level + trust_score), and enforce --require-cross so cross-vendor parity gates the verdict. Canonical body: ~/dev/mt-olympus/.claude/skills/cross-vendor-trust-gate/SKILL.md.

Mixed-model (cross-family) panel

When the decision wants a mixed-model / cross-family panel rather than single-model judges, the duel substrate is `/dual-pane-atm` — durable Opus + Codex (+AGY) panes, or its in-session variant (Agent model: override + codex exec) that `/discovery`'s fanout gate runs. /reverse-engineer routes one-way-door steal decisions into council.

Examples

/council should we swap the policy engine to Cedar? — runs at the canonical

location; this stub forwards. Read ~/dev/mt-olympus/.claude/skills/council/SKILL.md.

Troubleshooting

Skill seems empty / missing scripts: the body moved to Mount Olympus

(2026-06-10). Use the canonical path above; this stub exists only to keep fleet routing alive until the catalog closer updates the registry.

Runtime Contract

Multi-judge runs still bind to the shared Claude runtime surface: claude-code-latest-features.md.

{
  "type": "object",
  "properties": {
    "verdict": {"type": "string", "enum": ["PASS", "WARN", "FAIL"]},
    "confidence": {"type": "string", "enum": ["HIGH", "MEDIUM", "LOW"]},
    "key_insight": {"type": "string"},
    "findings": {
      "type": "array",
      "items": {
        "type": "object",
        "properties": {
          "severity": {"type": "string", "enum": ["critical", "significant", "minor"]},
          "category": {"type": "string"},
          "description": {"type": "string"},
          "location": {"type": "string"},
          "recommendation": {"type": "string"},
          "fix": {"type": ["string", "null"], "description": "Specific action to resolve this finding"},
          "why": {"type": ["string", "null"], "description": "Root cause or rationale"},
          "ref": {"type": ["string", "null"], "description": "File path, spec anchor, or doc reference"},
          "id": {"type": ["string", "null"], "description": "Optional finding identifier for cross-skill correlation (e.g., f-2026-03-12-001)"}
        },
        "required": ["severity", "category", "description", "location", "recommendation", "fix", "why", "ref", "id"],
        "additionalProperties": false
      }
    },
    "recommendation": {"type": "string"},
    "schema_version": {"type": "integer", "enum": [1, 2, 3, 4]},
    "satisfaction_score": {"type": ["number", "null"], "minimum": 0, "maximum": 1, "description": "Probabilistic satisfaction score (0.0 = unsatisfied, 1.0 = fully satisfied). Null means not computed."},
    "satisfaction_breakdown": {"type": ["object", "null"], "description": "Per-dimension satisfaction scores. Keys are dimension names, values are 0.0-1.0 scores. Null means not computed.", "additionalProperties": {"type": "number", "minimum": 0, "maximum": 1}}
  },
  "required": ["verdict", "confidence", "key_insight", "findings", "recommendation", "schema_version", "satisfaction_score", "satisfaction_breakdown"],
  "additionalProperties": false
}