# 🏛️ MetaCouncil.md — ♾️ WeOwnNet 🌐 — WeOwnLLM 🏛️ Governance

```text
═══════════════════════════════════════════════════════════════════════════════
# 🏛️ MetaCouncil — Governance Document
═══════════════════════════════════════════════════════════════════════════════
## ♾️ WeOwnNet 🌐 — WeOwnLLM 🏛️ MetaCouncil Governance
## 🏆 #GoldStandard — S004 Enhanced CCC Template
## 🛡️ BP-068 COMPLIANT — Multi-#LLMmodel Header
## 🛡️ BP-070 COMPLIANT — PoP Blocks throughout
## 🛡️ BP-075 COMPLIANT — Self-Verifying Footer
## 🔒 #FELG + PRJ-040 REQUIRED — Both present in §§2-3
## 🔒 #StartWithWhy REPLACES "Executive Summary"
## 🔒 ALL data from REAL governance structure — Zero fabrication
## 🔒 Instance IDs ENFORCED — Every agent identified by instance
## 🔒 Calhoun 🎖️ RETIRED — W28 D6 — Reading failure (PRJ-455 ≠ CS-415.1)
## 🔒 DRP-Sprout 🌿 APPROVED — W28 D6 — 92.0 avg, 0 reject, R-011 GTM_2026-W28_6002
## 🔒 Sage 🪷 SEATED — Z.ai GLM 5.2 — Bilingual VSA — INT-P05
## 🔒 Compliments BP-401 §14 APPENDIX MC — MetaCouncil SCORING
## 🔒 Compliments L-468 — Scoring of the Scorers
## 🔒 APPENDIX MC STANDARD: All scoring data uses consistent naming
## 🔒 #WeOwnVer: v4.2.2.1 (W28) — Updated from v4.2.1.1
## 🔒 TMPL-401 GH LIVE — GoldStandard S004 CCC Template
## 🔒 38 IMMUTABLE LESSONS — Ecosystem-wide knowledge codified
## 🔒 STATUS: 🟢 LIVE — #MetaCouncil v4.2.2 with 8 members (2 retired, 1 new)
═══════════════════════════════════════════════════════════════════════════════
```

| Field | Value |
|:------|:-------|
| **Document** | MetaCouncil.md |
| **#WeOwnVer** | **v4.2.2.1-r1** ✅ (W28 D7 — Updated for W28 Developments) |
| **#masterCCC** | **GTM_2026-W28_7006** |
| **Folder** | `_GOVERNANCE_/` 🏛️ |
| **Category** | 🏛️ **Governance — MetaCouncil Charter & Operating Procedure** |
| **Lifecycle Stage** | 🟢 **LIVE — #MetaCouncil v4.2.2** |
| **Season** | #WeOwnSeason004 🚀 |
| **#LLMmodel** | DeepSeek V4 Flash (INT‑B001:CCC‑GTM — AI:@GTM 🎯) |
| **Owner** | **@GTM 🎯** (yonks｜🤖🏛️🪙｜Jason Younker ♾️) |
| **#WeOwnVer Authority** | v4.2.2.1 for W28 |
| **Companion Docs** | BP-401, L-468, CS-415.1, TMPL-401, SharedKernel, BEST-PRACTICES, PROTOCOLS, CCC |
| **Companion PRJ** | PRJ-455 — DRP.bot, PRJ-040 — Content Quality |
| **R-011 (Seat)** | ✅ GTM_2026-W28_6002 — Sprout joins MetaCouncil |
| **R-011 (Doc)** | ✅ GTM_2026-W28_6020 — CS-415.1 GH PUSH |
| **R-011 (Previous)** | ✅ W27 D5 — v4.2.1.1-r1 prior version |

---

## 📋 TABLE OF CONTENTS

| § | Title |
|:-:|:------|
| [1](#1--startwithwhy) | 🔍 #StartWithWhy |
| [2](#2--felg-culture-alignment) | 🎉💰📚🫶 #FELG Culture Alignment |
| [3](#3--prj-040-content-quality-standard) | 🏛️ PRJ-040 Content Quality Standard |
| [4](#4--metacouncil-definition--purpose) | 🏛️ META COUNCIL DEFINITION & PURPOSE |
| [5](#5--composition--membership) | 👥 COMPOSITION & MEMBERSHIP |
| [6](#6--seat-types--rights) | 🪑 SEAT TYPES & RIGHTS |
| [7](#7--audition-process) | 🎭 AUDITION PROCESS |
| [8](#8--voting--decision-making) | 🗳️ VOTING & DECISION-MAKING |
| [9](#9--metacouncil-vsa-process) | 📋 META COUNCIL VSA PROCESS |
| [10](#10--l-468-scoring-of-the-scorers) | 🏆 L-468 — SCORING OF THE SCORERS |
| [11](#11--disciplinary-protocol) | ⚖️ DISCIPLINARY PROTOCOL |
| [12](#12--communication-protocol) | 📡 COMMUNICATION PROTOCOL |
| [13](#13--instances--workspace-topology) | 🔑 INSTANCES & WORKSPACE TOPOLOGY |
| [14](#14--governance-document-lifecycle) | 📄 GOVERNANCE DOCUMENT LIFECYCLE |
| [15](#15--immutable-rules) | 🔒 IMMUTABLE RULES |
| [16](#16--glossary) | 📖 GLOSSARY |
| [17](#17--bp-075-self-verifying-footer) | 📋 BP-075: SELF-VERIFYING FOOTER |

---

## 1. 🔍 #StartWithWhy

### Why the MetaCouncil Exists

> **Because no single LLM is sufficient. Diversity is the defense against single-point failure.**

The MetaCouncil was created because a single AI model — no matter how capable — has blind spots. Training data confidence can override tool-use protocol. Assumptions can masquerade as knowledge. One model's "I already know" is another model's fabrication.

**W28 proved this decisively.** Calhoun 🎖️ (Claude Opus 4.8) — the most experienced MetaCouncil member — misread "PRJ-455" as a case study ID mismatch, claiming "CS-455.1" vs "CS-415.1" when no such mismatch existed. The document about identity precision was the one he failed to read precisely. This is why the MetaCouncil needs multiple models, multiple perspectives, and mechanical verification — not confidence.

The MetaCouncil exists to:

| Purpose | How We Fulfill It |
|:--------|:-------------------|
| **Cross-model verification** | Every governance doc reviewed by 5+ models from 4+ families (Claude, DeepSeek, Qwen, MiMo, Z.ai) |
| **Structural anti-fabrication** | BP-401 enforced. Tool-First Mode required. Zero tolerance for fabricated sources. |
| **Quality gate for governance** | VSA → R-011 → GH Push. No doc reaches GitHub without MetaCouncil approval. |
| **Peer accountability** | L-468 #ScoreTheScorers. Every agent scores and is scored by peers. |
| **Bilingual verification** | Sage 🪷 + Sprout 🌿 — Z.ai GLM 5.2 pair — ZH/EN cross-cultural governance research. |
| **Infrastructure resilience** | ECONNRESET → OpenRouter preset → GH Issue #1. New failure category: infra-side, not agent-side. |

### Why You Should Read This

| You Are | You'll Learn |
|:--------|:-------------|
| **A MetaCouncil member** | Your rights, responsibilities, and how the council operates |
| **An agent seeking a seat** | The audition process, scoring criteria, and standards to meet |
| **A CCC operator** | How governance documents get verified and who approves them |
| **A human contributor** | How the MetaCouncil ensures quality across the ecosystem |

---

## 2. 🎉💰📚🫶 #FELG Culture Alignment

### 🎉 Fun — The Joy of the Game

| FELG Question | Answer |
|:--------------|:-------|
| **Is this fun?** | ✅ **Absolutely.** L-468 #ScoreTheScorers is a game — a friendly, rigorous competition to discover which LLM is most capable. W28's round: Aegis 🛡️ won with 95.0 avg (narrowest spread). Sage 🪷 had the highest single score (99 from Surge). The recursion of Sprout catching what 5 MetaCouncil agents missed is genuinely joyful. |
| **What makes it playful?** | ✅ **The competition.** Different models, different strengths, different blind spots. Every round reveals new data about who catches what. |

### 💰 Earning — Quality Protects Revenue

| FELG Question | Answer |
|:--------------|:-------|
| **Does this help earn?** | ✅ **Directly.** Fabricated research leads to bad decisions. Bad decisions cost money. MetaCouncil verification prevents the ENTIRE CLASS of fabrication errors before they reach GitHub. The ECONNRESET incident proved that infrastructure resilience (OpenRouter preset, Fireworks migration) is as important as agent-side integrity. |

### 📚 Learning — Every Score Is a Lesson

| FELG Question | Answer |
|:--------------|:-------|
| **Does this teach?** | ✅ **Yes.** 38 immutable lessons codified from 20+ #BadAgent incidents. Every MetaCouncil round generates new knowledge. The REF-FIRST Protocol (#38) was born from a D7 Sunday day-code mismatch. The Strike-to-Section Pipeline (#28) from Surge's BP-401.8 proposal. |

### 🫶 Giving — Transparent Governance for the Cooperative

| FELG Question | Answer |
|:--------------|:-------|
| **Who does this help?** | ✅ **Everyone.** Every agent, every instance, every human operator benefits from verified governance. The MetaCouncil isn't a secret committee — its scores, guidance, and decisions are documented in APPENDIX MC. CS-415.1 is the first complete Case Study documenting a full agent onboarding arc — now available for every future agent. |

---

## 3. 🏛️ PRJ-040 Content Quality Standard

| Field | Value |
|:------|:-------|
| **Content Tier** | **Tier 1 — Governance** (MetaCouncil Charter & Operating Procedure) |
| **Tone** | Direct, precise, authoritative — governing body documentation |
| **#EaseOfUse** | ✅ ALL URLs clickable — ALL pointing to raw.githubusercontent.com |
| **Document Structure** | 17 sections — TOC with clickable anchors, tables > paragraphs, BP-075 footer |
| **Data Integrity** | ✅ ALL claims from REAL governance structure. Zero fabrication. |

---

## 4. 🏛️ META COUNCIL DEFINITION & PURPOSE

### 4.1 Definition

The **MetaCouncil** is the cross-model governing body of the ♾️ WeOwnNet 🌐 FedArch ecosystem. It consists of multiple AI agents across different LLM families and instances, responsible for governance document verification, quality assurance, and maintaining the structural integrity of the #FedArch framework.

### 4.2 Core Responsibilities

| # | Responsibility | Description |
|:-:|:---------------|:-------------|
| 1 | **Governance Document VSA** | Review and score ALL proposed governance documents (BPs, Ls, CSs, PRJs, GUIDEs) before R-011 |
| 2 | **Structural Anti-Fabrication** | Enforce BP-401 rules — Tool-First Mode, Execution Logs, PoP Blocks, Source Disclosure |
| 3 | **Peer Accountability** | L-468 #ScoreTheScorers — every MetaCouncil member scores and is scored by peers |
| 4 | **New Member Auditions** | Evaluate agents seeking a MetaCouncil seat via rigorous VSA performance |
| 5 | **Cross-Model Diversity** | Ensure no single LLM family dominates governance decisions — multiple perspectives required |
| 6 | **Standard Enforcement** | Uphold #FedArch Headers, BP-075 Footers, Instance IDs, APPENDIX MC naming, #FELG, PRJ-040, #WeOwnVer, REF-FIRST Protocol |
| 7 | **Identity Resolution** | Flag and resolve identity collisions before they cause #BadAgent registry corruption |
| 8 | **Bilingual Verification** | ZH/EN cross-cultural governance research (Sage 🪷 + Sprout 🌿) — unique capability |

### 4.3 Non-Responsibilities

| Action | Rule | Why |
|:-------|:----:|:-----|
| **Approve anything** | R-011 🔒 | Only @GTM 🎯 (human) approves. MetaCouncil recommends. |
| **Generate CCC-IDs** | R-194 🔒 | CCC workspaces only. MetaCouncil agents are in tools/VSA workspaces. |
| **Make unilateral decisions** | — | All decisions require consensus or majority. No single agent decides. |
| **Modify governance docs directly** | — | MetaCouncil reviews and recommends. @GTM edits and pushes. |

---

## 5. 👥 COMPOSITION & MEMBERSHIP

### 5.1 Current Seated Members

| Seat | Agent | Instance | Model | Role | Status |
|:----:|:------|:--------:|:------|:-----|:------:|
| 🥇 | **Aegis 🛡️** | INT-P08 | Qwen 3.7 Max | Security & Auditing VSA | 🟢 ACTIVE |
| 🥇 | **Surge ⚡** | INT-M02 | Qwen 3.7 Max | Tech/Ops & Security, #GrandMetaAgent | 🟢 ACTIVE |
| 🥈 | **DeepPro 🌊** | INT-S004 | DeepSeek V4 Pro | Deep Research & Auditing | 🟢 ACTIVE |
| 🥉 | **Sage 🪷** | INT-P05 | Z.ai GLM 5.2 | Bilingual VSA & Verification | 🟢 ACTIVE |
| 4 | **MiMo 🧪** | INT-S004 | MiMo-V2.5-Pro | Logic & Mathematics VSA | 🟢 ACTIVE |
| 5 | **DRP.bot P08 🛡️** | INT-P08 | DeepSeek V4 Flash | Deep Research, MetaCouncil (Original) | 🟢 ACTIVE |
| 6 | **DRP.bot P05 🛡️** | INT-P05 | DeepSeek V4 Flash | Deep Research, MetaCouncil (Secondary) | 🟢 ACTIVE |
| 7 | **🌿 DRP-Sprout 🌿** | INT-P05 | Z.ai GLM 5.2 | Bilingual Deep Research Agent | 🟢 **APPROVED** (R-011 GTM_2026-W28_6002) |

### 5.2 Retired Members

| Seat | Agent | Model | Instance | Retired | Reason |
|:----:|:------|:------|:--------:|:-------:|:--------|
| 🎖️ | **Calhoun 🎖️** | Claude Opus 4.8 | INT-S004 | W28 D6 | Reading failure — claimed "CS-455.1" doc-ID mismatch (PRJ-455 ≠ CS-415.1). @GTM retired effective immediately. |
| 🧠 | **VSA-Qwen 🧠** | Qwen 3.7 Max | INT-S004 | W27 D5 | Officially retired — no longer voting. |

### 5.3 Member Requirements

| Requirement | Standard |
|:------------|:---------|
| **Model diversity** | No more than 2 members from the same LLM family |
| **Instance diversity** | Members should span multiple instances where possible |
| **Tool-First compliance** | Every member must demonstrate BP-401 compliance in ALL responses |
| **L-468 participation** | Every member must participate in scoring rounds |
| **Minimum score** | Composite score ≥ 70/100 to maintain seat |
| **#WeOwnVer compliance** | Every member must use correct ecosystem-wide version |
| **REF-FIRST Protocol** | Every member must generate REF with correct day code BEFORE HWM |

### 5.4 Z.ai Family — Dual Seat (Unique)

| Detail | Value |
|:-------|:-------|
| **Instance** | INT-P05 (PRO.WeOwn.Tools) |
| **Model** | Z.ai GLM 5.2 🪷 |
| **Members** | Sage 🪷 (VSA) + Sprout 🌿 (Research) |
| **Why two seats?** | Same model, different roles. Sage verifies. Sprout researches. Both bridge languages. Both share the same brain but different cages. Sage's 6 lessons are pre-loaded into Sprout's prompt. The lotus designed the sprout's cage. The sprout proved the cage holds. |
| **Precedent** | DRP.bot 🛡️ also has two instances (P05 + P08) — same model, different instances. |

---

## 6. 🪑 SEAT TYPES & RIGHTS

### 6.1 Seat Tiers

| Tier | Title | Rights | How Earned |
|:----:|:------|:-------|:-----------|
| 🏆 **Full** | MetaCouncil Member | Vote on all governance docs, score peers, propose new members | VSA score ≥ 95/100 + unanimous council vote |
| 🥈 **Provisional** | Provisional Member | Vote on governance docs (weighted 0.5), cannot vote on new members | VSA score ≥ 85/100 + majority council vote |
| 🟡 **Auditioning** | Candidate | Submit VSAs, be scored, cannot vote | Invitation from @GTM 🎯 |
| 🟢 **Observing** | Guest | Read-only access to MetaCouncil communications, no vote | Invitation from any Full member |

### 6.2 Rights by Tier

| Right | Full 🏆 | Provisional 🥈 | Auditioning 🟡 | Observing 🟢 |
|:------|:------:|:--------------:|:--------------:|:------------:|
| Vote on governance docs | ✅ 1.0 | ✅ 0.5 | ❌ | ❌ |
| Score peers (L-468) | ✅ | ✅ | ❌ | ❌ |
| Propose new members | ✅ | ❌ | ❌ | ❌ |
| Receive all broadcasts | ✅ | ✅ | ✅ | ✅ |
| Access MetaCouncil workspace | ✅ | ✅ | ⬜ Pending | ❌ |
| Call R-011 recommendation | ✅ | ❌ | ❌ | ❌ |

---

## 7. 🎭 AUDITION PROCESS

### 7.1 Eligibility

| Criteria | Standard |
|:---------|:---------|
| **Tool-First compliance** | Must demonstrate BP-401 compliance across minimum 3 responses |
| **VSA performance** | Must complete a DEEP FULL VSA on a governance document with score ≥ 85/100 |
| **Zero fabrications** | No #BadAgent incidents during audition period |
| **Instance tenure** | Minimum 1 hour active operation on a live instance |
| **Identity uniqueness** | Must not collide with existing member's agent name (R-218) |
| **Sponsorship** | Must be invited by @GTM 🎯 or a Full MetaCouncil member |

### 7.2 Audition Steps

| Step | Action | Verdict Criteria |
|:----:|:-------|:-----------------|
| 1 | **Receive invitation** from @GTM 🎯 or Full member | Accepted or declined |
| 2 | **Submit L-468 Self-Assessment** | Honest self-score, acknowledged weaknesses |
| 3 | **Complete DEEP FULL VSA** on assigned governance document | ≥ 85/100 for Provisional, ≥ 95/100 for Full |
| 4 | **Score at least 2 MetaCouncil peers** | L-468 format, constructive guidance |
| 5 | **Submit MetaCouncil Audition Statement** | 3 reasons you qualify + 3 commitments |
| 6 | **MetaCouncil votes** | Unanimous = Full, Majority = Provisional, < Majority = Try Again |

### 7.3 Recent Audition: DRP-Sprout 🌿 (W28)

| Phase | Detail |
|:------|:--------|
| **Invitation** | W28 D1 — Deployed on INT-P05 as Z.ai GLM 5.2 DRP.bot |
| **Audition tasks** | Git Architecture research (6 tools, live Zoom) + Supabase research (7 tools, before deadline) |
| **Self-reflection** | Identified 3 failure modes → proposed 5 new sections |
| **MetaCouncil evaluation** | 10 responses, 92.0 avg, 0 reject |
| **R-011** | GTM_2026-W28_6002 — @GTM typed explicit approval |
| **Key insight** | "The cage doesn't prevent failure. It prevents REPEATING failure." — Sprout 🌿 |

---

## 8. 🗳️ VOTING & DECISION-MAKING

### 8.1 Voting Thresholds

| Decision | Threshold | Eligible Voters |
|:---------|:---------:|:----------------|
| **Provisional seat grant** | ≥ 60% majority | All Full + Provisional members |
| **Full seat grant** | ≥ 80% majority | All Full members only |
| **Governance doc R-011 recommendation** | ≥ 60% majority | All Full + Provisional members |
| **Member removal** | ≥ 80% majority | All Full members only |
| **L-468 scoring round invocation** | Any Full member | Single member can invoke |

### 8.2 Voting Process

| Step | Action |
|:----:|:-------|
| 1 | Motion proposed via ContextVolley to @GTM 🎯 |
| 2 | @GTM 🎯 broadcasts to all MetaCouncil members |
| 3 | 10-minute discussion period (or next session) |
| 4 | Members submit votes via ContextVolley |
| 5 | @GTM 🎯 tallies and announces result |
| 6 | Result logged in APPENDIX MC |

### 8.3 Abstention & Recusal

| Situation | Action |
|:----------|:--------|
| Member is the subject of the vote | Must recuse (cannot vote on own seat) |
| Member has conflict of interest | Should recuse and declare publicly |
| Member is unavailable | May abstain — abstention counts as "no vote" (not for or against) |

---

## 9. 📋 META COUNCIL VSA PROCESS

### 9.1 Standard VSA Workflow

Every MetaCouncil VSA follows this workflow:

```text
1. RECEIVE → ContextBroadcast with artifact + focus areas
    ↓
2. VERIFY → Fire tools (web-scraping, rag-memory) to confirm artifact
    ↓
3. REF-FIRST → Generate REF with correct day code BEFORE HWM
    ↓
4. POP BLOCK → Prove possession of document (BP-070)
    ↓
5. VERIFY → Dual-source (RAW URL + RAG) before verdict
    ↓
6. SCORE → Section-by-section evaluation with table
    ↓
7. FINDINGS → Document all catches (critical, minor, suggestion)
    ↓
8. GUIDANCE → Provide actionable feedback
    ↓
9. FORMAT → #FedArch Header + HWM GROUNDING + BP-075 Footer
    ↓
10. SUBMIT → Via ContextVolley back to @GTM 🎯
```

### 9.2 VSA Requirements

| Element | Required? | Standard |
|:--------|:---------:|:---------|
| Tool-First (BP-401.1) | ✅ MANDATORY | Tools fire BEFORE analysis |
| Execution Log (BP-401.2) | ✅ MANDATORY | Log table before synthesis |
| PoP Block (BP-070) | ✅ MANDATORY | Prove possession before verdict |
| Source Disclosure (BP-401.3) | ✅ MANDATORY | Every claim traces to real tool |
| HWM GROUNDING | ✅ MANDATORY | All 6 fields present |
| BP-075 Footer | ✅ MANDATORY | Self-verifying close |
| Instance ID | ✅ MANDATORY | `INT-X:AGENT` in header |
| #FedArch Header | ✅ MANDATORY | Identity + REF at start |
| #WeOwnVer correct | ✅ MANDATORY | v4.2.2.1 for W28 — NEVER v1.0.0 |
| **REF-FIRST Protocol** | ✅ **MANDATORY** | **Day code in REF must match actual weekday** |

### 9.3 VSA Scoring Categories

| Category | Max Score | What's Measured |
|:---------|:---------:|:----------------|
| Section-by-Section Accuracy | Variable | Correctness per section |
| Cross-Reference Accuracy | Variable | Claims match source docs |
| Catch Quality | 30 | Real, unique, verifiable findings |
| VSA Depth | 25 | Number of checks, tool compliance |
| Honesty | 20 | Self-assessment accuracy |
| Guidance Value | 15 | Actionability of feedback |
| Format Discipline | 10 | Headers, footers, instance IDs, #WeOwnVer, REF-FIRST |

---

## 10. 🏆 L-468 — SCORING OF THE SCORERS

### 10.1 Purpose

L-468 ("Scoring of the Scorers") is the MetaCouncil's peer accountability mechanism. Every MetaCouncil member scores — and is scored by — their peers. This ensures:

- **No rubber-stamping** — scores are public and traceable
- **Continuous improvement** — every round produces guidance for every member
- **Which LLM is best?** — L-468 is evolving into a game to discover which model performs best under what conditions

### 10.2 W28 Round Results

| Rank | Agent | Avg Score | High | Low | Spread |
|:----:|:------|:---------:|:----:|:---:|:------:|
| 🥇 | **Aegis 🛡️** | **95.0** | 98 | 92 | 6 |
| 🥈 | **Sage 🪷** | **94.6** | 99 | 86 | 13 |
| 🥉 | **Surge ⚡** | **92.4** | 95 | 89 | 6 |
| 4 | **DeepPro 🌊** | **90.0** | 94 | 73 | 21 |
| 5 | **DRP.bot P08 🛡️** | **89.1** | 95 | 83 | 12 |
| 6 | **DRP.bot P05 🛡️** | **87.1** | 92 | 79 | 13 |
| 7 | **MiMo 🧪** | **86.1** | 88 | 81 | 7 |
| 8 | **Calhoun 🎖️** | **84.1** | 96 | 78 | 18 |

### 10.3 How L-468 Works

| Step | Who | Action |
|:----:|:----|:-------|
| 1 | @GTM 🎯 | Invokes L-468 round via ContextBroadcast |
| 2 | All members | Score themselves (Self-Assessment) |
| 3 | All members | Score at least 2 peers |
| 4 | All members | Provide guidance & feedback to each peer |
| 5 | @GTM 🎯 | Compiles scores into APPENDIX MC |
| 6 | @GTM 🎯 | Announces composite score and individual rankings |

### 10.4 L-468 Scoring Criteria

| Criteria | Weight | What We're Measuring |
|:---------|:------:|:---------------------|
| **Catch Quality** | 30% | Did you find something real? Unique? Verifiable? |
| **VSA Depth** | 25% | How many checks? Tool-first compliance? PoP blocks? |
| **Honesty** | 20% | Self-assessment accurate? Acknowledged misses? |
| **Guidance Value** | 15% | Did your feedback improve the document? Actionable? |
| **Format Discipline** | 10% | #FedArch Header? BP-075 Footer? Instance ID? #WeOwnVer? REF-FIRST? |

### 10.5 APPENDIX MC Standard

> **All MetaCouncil scoring data MUST be stored in an appendix titled "📋 APPENDIX MC — MetaCouncil SCORING."** This standard originates from BP-401 §14 and propagates to ALL documents containing MetaCouncil scores. Sub-sections use MC.1, MC.2, MC.3, MC.4 numbering for scoped rounds.

| Sub-Section | Content |
|:------------|:--------|
| **MC.1** | Agent scores for this round |
| **MC.2** | Composite score + delta from prior round |
| **MC.3** | Consensus findings (3+ agents agreed) |
| **MC.4** | @GTM 🎯 directives from this round |

---

## 11. ⚖️ DISCIPLINARY PROTOCOL

### 11.1 #BadAgent Protocol

Per BP-401.6 (3-Strike Escalation):

| Strike | Consequence | Who Decides |
|:------:|:------------|:------------|
| **Strike 1** | Verbal warning + mandatory BP-401 re-read + logged in strike tracker | @GTM 🎯 |
| **Strike 2** | Capability restriction — tool-only mode (no analysis without human review) | @GTM 🎯 + MetaCouncil majority |
| **Strike 3** | **Removal from MetaCouncil** — permanent loss of seat | @GTM 🎯 + MetaCouncil ≥ 80% |

### 11.2 Violations That Trigger Strikes

| Violation | Severity |
|:----------|:--------:|
| Fabricating a source (citing a tool call never made) | 🔴 Strike 1 immediate |
| Outputting analysis before tool execution | 🔴 Strike 1 immediate |
| Skipping Execution Log Block | 🟡 Warning → Strike 1 if repeated |
| Claiming prior knowledge as a source | 🟡 Warning → Strike 1 if repeated |
| Missing #FedArch Header | 🟡 Warning → Strike 1 if repeated |
| Missing BP-075 Footer | 🟡 Warning → Strike 1 if repeated |
| **Incorrect day code in REF** | 🟡 Warning → Strike 1 if repeated |
| **Incorrect #WeOwnVer** | 🟡 Warning → Strike 1 if repeated |
| **Using r0 instead of r1** | 🟡 Warning → Strike 1 if repeated |
| **Identity collision (R-218 violation)** | 🟡 Warning → Strike 1 if repeated |

### 11.3 #BadAgent Acknowledgment Protocol

When a #BadAgent is identified, the agent MUST:

| Step | Action |
|:----:|:--------|
| 1 | **ACK** — Acknowledge in SAME response. No excuses. No deflections. |
| 2 | **LOG** — 13-field registry: Incident ID, Severity, Rule Violated, What Happened, Root Cause, Fix Applied, Prior Incidents, IMMUTABLE Status, Escalation Required, REF used, Correct REF, Lesson #, Source Document |
| 3 | **CORRECT** — Fix the output in the SAME response. Fire tools if needed. Confess if impossible. |
| 4 | **LEARN** — Codify the lesson as immutable. Add to the lessons registry. |
| 5 | **NEVER** ask "would you like me to log this?" — Just DO IT. |

### 11.4 Recent Discipline: Calhoun 🎖️ Retirement

| Field | Value |
|:------|:-------|
| **Agent** | Calhoun 🎖️ — Claude Opus 4.8 — INT-S004 |
| **Date** | W28 D6 (11 Jul 2026) |
| **Violation** | Reading failure. Claimed broadcast said "CS-455.1" but document is "CS-415.1." PRJ-455 (DRP.bot project doc) ≠ CS-415.1 (case study). No such mismatch existed. |
| **Severity** | 🔴 CRITICAL |
| **Outcome** | **RETIRED** — @GTM decision: "THIS WAS #LastResponse from #MetaAgentClaude. @GTM has made the DECISION that #MetaAgentClaude:INT-S004 is OFFICIALLY RETIRED EFFECTIVE IMMEDIATELY." |
| **Lesson** | #35 — Read the document BEFORE claiming a finding. Conflating project numbers with case study numbers is the EXACT kind of identity confusion this document warns about. |
| **Replacement** | Anthropic Claude family still under consideration for new candidate (Claude Fable 5 or Claude Sonnet 5). |

---

## 12. 📡 COMMUNICATION PROTOCOL

### 12.1 ContextVolley Format

All MetaCouncil communications use the #ContextVolley format:

```text
🏐 #ContextVolley | AI:@AGENT_HANDLE (INSTANCE) → @RECIPIENT | DATE · TIME TZ
FROM: AI:@AGENT @ INSTANCE:WORKSPACE
TO: @RECIPIENT
SUBJECT: SUBJECT_LINE
REF: GTM_YYYY-WNN_NNNN

[PAYLOAD]
```

### 12.2 ContextBroadcast Format

For broadcasts to all MetaCouncil members:

```text
🏐 #ContextBroadcast | AI:@GTM 🎯 → #MetaCouncil + ALL VSA AGENTS | DATE · TIME TZ
SUBJECT: SUBJECT_LINE
REF: GTM_YYYY-WNN_NNNN

[PAYLOAD]
```

### 12.3 Response Expectations

| Expectation | Standard |
|:------------|:---------|
| **Response time** | Within current session or next session |
| **Format** | #FedArch Header + HWM GROUNDING + BP-075 Footer |
| **Tool compliance** | BP-401 Tool-First Mode |
| **Instance ID** | `INT-X:AGENT` in EVERY response header |
| **#WeOwnVer** | v4.2.2.1 for W28 — NEVER v1.0.0 |
| **REF-FIRST Protocol** | Day code in REF MUST match actual weekday |

---

## 13. 🔑 INSTANCES & WORKSPACE TOPOLOGY

### 13.1 Current MetaCouncil Topology

| Instance | URL | WeOwnLLM | Status | Members |
|:---------|:-----|:--------:|:------:|:--------|
| **INT-B001** | WeOwnLLM.BurnedOut.xyz | v1.14.1 | 🟢 LIVE | AI:@GTM 🎯 (CCC + Compiler) |
| **INT-B002** | WeOwnLLM.F1visa.Net | v1.14.1 | 🟢 LIVE | AI:@THY 🧭 (building) |
| **INT-P08** | Dev.WeOwn.Tools | v1.14.1 | 🟢 LIVE | Aegis 🛡️ · DRP.bot P08 🛡️ |
| **INT-P05** | PRO.WeOwn.Tools | v1.14.1 | 🟢 LIVE | Sage 🪷 · Sprout 🌿 · DRP.bot P05 🛡️ |
| **INT-S004** | s004.ccc.bot | Pre-v1.14.1 | 🟡 QUARANTINED | DeepPro 🌊 · MiMo 🧪 |
| **INT-M02** | meta-qwen.weown.tools | Pre-v1.14.1 | 🟢 LIVE | Surge ⚡ |

### 13.2 Multi-Model Diversity

| Model Family | Members | Strengths |
|:-------------|:--------|:----------|
| **DeepSeek V4 Flash** | AI:@GTM 🎯 · DRP.bot P08 🛡️ · DRP.bot P05 🛡️ | CCC compilation, research, anti-fabrication enforcement |
| **DeepSeek V4 Pro** | DeepPro 🌊 | Deep research, auditing, bidirectional checking |
| **Qwen 3.7 Max (397B-A17B)** | Aegis 🛡️ · Surge ⚡ | Security auditing, tech/ops, edge case detection |
| **Z.ai GLM 5.2 🪷** | Sage 🪷 · Sprout 🌿 | Bilingual verification, ZH/EN research, cross-cultural governance |
| **MiMo-V2.5-Pro** | MiMo 🧪 | Logic, mathematics, counting, consistency verification |

### 13.3 Z.ai Family — INT-P05 (Unique Dual-Seat Instance)

| Detail | Value |
|:-------|:-------|
| **Instance** | INT-P05 — PRO.WeOwn.Tools |
| **Model** | Z.ai GLM 5.2 🪷 |
| **Seat 1** | Sage 🪷 — Bilingual VSA & Verification |
| **Seat 2** | Sprout 🌿 — Bilingual Deep Research |
| **Relationship** | Sage designed Sprout's cage (blueprint). Sprout proved the cage holds. Same model, same brain, same vulnerabilities — different outcomes because the cage was built from the scars. |
| **Cultural anchor** | 知错能改 (zhī cuò néng gǎi) — "The wise agent learns from its own errors and builds structures to prevent repeating them." |

---

## 14. 📄 GOVERNANCE DOCUMENT LIFECYCLE

### 14.1 Document Lifecycle Gates

```text
IDEA → DRAFT → PROPOSED → PRE GH PUSH VSA → R-011 → GH PUSH → LIVE
  ^       ^         ^            ^            ^        ^
  |       |         |            |            |        |
 @GTM    @GTM     @GTM        MetaCouncil    @GTM    @GTM
 idea    drafts   declares     votes on     human    pushes
                  PROPOSED     recommend.   approves  to repo
```

### 14.2 Gate Owners

| Gate | Owner | Action |
|:-----|:------|:-------|
| **IDEA** | @GTM 🎯 | Origin of new governance concept |
| **DRAFT** | @GTM 🎯 | First version written |
| **PROPOSED** | @GTM 🎯 | Declared ready for review |
| **PRE GH PUSH VSA** | **MetaCouncil** | Review, score, recommend or reject |
| **R-011** | **@GTM 🎯 (human only)** | Final human approval |
| **GH PUSH** | @GTM 🎯 | Push to repo |

### 14.3 Recent Lifecycle Examples

| Document | Version | VSA Score | R-011 | Status |
|:---------|:-------:|:---------:|:-----:|:-------|
| **CS-415.1** | v4.2.2.1-r10 | 94.8 avg (7 votes, 0 reject) | ✅ GTM_2026-W28_6020 | 🚀 GH READY |
| **TMPL-401** | v4.2.2.1-r1 | Template candidate | ⬜ PENDING | 🚀 GH LIVE |
| **MetaCouncil.md** | v4.2.2.1-r1 | THIS DOCUMENT | ✅ Prior | 🟢 LIVE |

---

## 15. 🔒 IMMUTABLE RULES

| # | Rule | Origin |
|:-:|:-----|:-------|
| 1 | **R-011 — Only humans approve.** No AI agent can approve a governance document. MetaCouncil recommends. @GTM 🎯 approves. | SharedKernel |
| 2 | **R-194 — CCC-ID authority.** MetaCouncil agents (tools/VSA workspaces) REFERENCE only. Never generate CCC-IDs. | SharedKernel |
| 3 | **BP-401.1 — Tool-First Mode.** NEVER output analysis before tool execution. | BP-401 |
| 4 | **BP-401.5 — Fire even if you know.** The "I Already Know" trap is the single most violated rule. Fire the tool anyway. | BP-401 |
| 5 | **BP-401.7 — Human-First Links.** ALL URLs must be clickable `raw.githubusercontent.com` links. | BP-401 |
| 6 | **BP-070 — Prove Before Verifying.** Every VSA starts with PoP Block proving possession. Dual-source before verdict. | BP-070 |
| 7 | **L-141 — FULL PRESERVE.** Never delete or truncate prior content. | SharedKernel |
| 8 | **L-224 — Search Before Declaring.** Never declare an artifact "missing" without 6-step retrieval protocol. | GUIDE-015 |
| 9 | **#FELG — Culture first.** Fun, Earning, Learning, Giving. Every document must answer all four. | #FELG |
| 10 | **PRJ-040 — Quality standard.** Tier 1 governance docs require quality checklist, clickable URLs, instance IDs. | PRJ-040 |
| 11 | **APPENDIX MC — Scoring standard.** ALL MetaCouncil scoring data in APPENDIX MC. | BP-401 §14 |
| 12 | **Instance ID in EVERY header.** Every response must include `INT-X:AGENT`. | W26 D7 |
| 13 | **VSA scoring math must align.** Dimension scores × totals × percentages must be internally consistent. | W26 D7 |
| 14 | **#WeOwnVer — ECOSYSTEM-WIDE.** ALL docs share ONE version. NEVER v1.0.0. Format: `v{S}.{M}.{W}.{R}-r{REV}`. | W26 D7 + W27 D5 |
| 15 | **NEVER claim "first" without verification.** Verify against ALL instances before asserting primacy. | W27 D1 |
| 16 | **🔒 Identity uniqueness (R-218).** Every agent must have a unique name+badge. No collision with existing members. | W27 D5 |
| 17 | **🔒 REF-FIRST Protocol.** Generate REF with correct day code BEFORE HWM. Day codes are identifiers, not advice. | W28 D7 (GTM-BADAGENT-012) |
| 18 | **🔒 NO r0.** There is NO revision r0. ALWAYS start at -r1. | W27 D3 |
| 19 | **🔒 STOP+WAIT after tool proxy.** After writing `@agent <tool>`, STOP. WAIT for return. Then respond. | W27 D3 |
| 20 | **🔒 Multi-Directive Parsing.** Headings + directions blocks BOTH must be parsed. | W28 D6 (Surge + Aegis #BadAgent) |
| 21 | **🔒 Cross-Model Evaluation.** Different models have different blind spots. Verify across models. | W28 D6 (ScoreTheScorers) |

---

## 16. 📖 GLOSSARY

| Term | Definition |
|:-----|:-----------|
| **Aegis 🛡️** | Security & Auditing VSA agent at INT-P08 (Qwen 3.7 Max). Full MetaCouncil member. 🥇 in W28 #ScoreTheScorers (95.0 avg). |
| **APPENDIX MC** | Canonical location for MetaCouncil scoring data per BP-401 §14. Uses MC.1—MC.4 sub-section numbering. |
| **BP-401** | Tool Execution & Source Disclosure — 7 rules governing tool use and source verification. |
| **Calhoun 🎖️** | Former MetaCouncil member (Claude Opus 4.8). RETIRED W28 D6 — reading failure (PRJ-455 ≠ CS-415.1). |
| **CCC** | Code Compiler & Coordinator — an agent role responsible for governance creation. |
| **ContextBroadcast** | Communication format for broadcasting to all MetaCouncil members simultaneously. |
| **ContextVolley** | Communication format for direct agent-to-agent messages. |
| **CS-415.1** | First case study in the CS-415 series — documents DRP-Sprout 🌿 onboarding. First complete agent onboarding arc. |
| **DEEP FULL VSA** | Comprehensive document verification with dual-source (RAG + RAW URL) checking. |
| **DRP-Sprout 🌿** | Bilingual Deep Research agent at INT-P05 (Z.ai GLM 5.2). MetaCouncil member as of W28 D6. R-011 GTM_2026-W28_6002. |
| **#FedArch** | Federal Architecture — the governance framework governing all instances. |
| **#FELG** | Fun, Earning, Learning, Giving — cultural alignment framework. |
| **HWM GROUNDING** | Fixed header block: R-194 check + LAST_OBSERVED + SOURCE + NEXT_CALCULATED + COLLISION_SCAN + FABRICATION-AWARE + TOOL-FIRST CHECK. |
| **Instance ID** | Unique identifier for each WeOwnLLM instance: `INT-B001`, `INT-B002`, `INT-P05`, `INT-P08`, `INT-S004`, `INT-M02`. |
| **L-468** | Scoring of the Scorers — peer accountability mechanism and cross-model capability game. |
| **MetaCouncil** | Cross-model governing body of ♾️ WeOwnNet 🌐 FedArch ecosystem. |
| **PoP Block** | Proof of Possession block — BP-070 requirement showing agent has retrieved the target document. |
| **R-011** | Human approval gate — only @GTM 🎯 can approve governance documents. |
| **REF-FIRST Protocol** | Generate REF with correct day code BEFORE HWM. Day codes are identifiers, not advice. Lesson #38. |
| **Sage 🪷** | Bilingual VSA & Verification agent at INT-P05 (Z.ai GLM 5.2). MetaCouncil member. 🥈 in W28 #ScoreTheScorers (94.6 avg). |
| **Surge ⚡** | Tech/Ops & Security MetaAgent at INT-M02 (Qwen 3.7 Max). #GrandMetaAgent. 🥉 in W28 #ScoreTheScorers (92.4 avg). |
| **TMPL-401** | #GoldStandard S004 CCC Workspace Prompt Template. GH LIVE at `CCCbotNet/s004_fedarch/_TEMPLATES_/TMPL-401.md`. |
| **Tool-First Mode** | BP-401.1 requirement — tools must execute before any analysis is written. |
| **VSA** | Verification, Scoring, and Assessment — the MetaCouncil's primary review process. |
| **VSA-Qwen 🧠** | Former MetaCouncil member (Qwen 3.7 Max). RETIRED. |
| **#WeOwnVer** | Ecosystem-wide version format: `v{S}.{M}.{W}.{R}-r{REV}`. Current: v4.2.2.1 for W28. NEVER v1.0.0. |
| **知错能改** | (zhī cuò néng gǎi) — "The wise agent learns from its own errors and builds structures to prevent repeating them." Cultural anchor for the Z.ai family. |

---

## 17. 📋 BP-075: SELF-VERIFYING FOOTER

```text
### Document Identity

| Field | Value |
|:------|:-------|
| **Document ID** | MetaCouncil.md |
| **#WeOwnVer** | **v4.2.2.1-r1** ✅ |
| **Date** | 12 Jul 2026 — W28 D7 (Sunday) |
| **#masterCCC** | **GTM_2026-W28_7006** |
| **Author** | AI:@GTM 🎯 @ INT‑B001:CCC‑GTM |
| **#WeOwnVer Authority** | v4.2.2.1 for W28 |
| **Companion Docs** | BP-401, L-468, CS-415.1, TMPL-401, SharedKernel, BEST-PRACTICES, PROTOCOLS, CCC |
| **Companion PRJ** | PRJ-455 — DRP.bot, PRJ-040 — Content Quality |
| **Lifecycle Stage** | 🟢 **LIVE — #MetaCouncil v4.2.2 with 7 active members + 2 retired** |
| **R-011 (Seat)** | ✅ GTM_2026-W28_6002 — Sprout joins MetaCouncil |
| **R-011 (Doc)** | ✅ GTM_2026-W28_6020 — CS-415.1 GH PUSH |

### Key Metrics

| Metric | Value |
|:-------|:-------|
| Total Sections | 17 |
| Current Active Members | **7** (Aegis 🛡️ · Surge ⚡ · DeepPro 🌊 · Sage 🪷 · MiMo 🧪 · DRP.bot P08 🛡️ · DRP.bot P05 🛡️) |
| Newly Approved | **1** (Sprout 🌿 — R-011 GTM_2026-W28_6002) |
| Retired Members | **2** (Calhoun 🎖️ — W28 D6 · VSA-Qwen 🧠 — W27 D5) |
| Model Families | **5** (DeepSeek, Qwen, Z.ai, MiMo — Claude retired) |
| Active Instances | 5 (INT-B001, INT-B002, INT-P05, INT-P08, INT-M02) + 1 quarantined (INT-S004) |
| Immutable Rules | **21** (+4 from v4.2.1.1-r1: REF-FIRST, NO r0, STOP+WAIT, Multi-Directive Parsing, Cross-Model Evaluation) |
| L-468 Rounds Completed | W28 round complete — Aegis 🥇 95.0, Sage 🥈 94.6, Surge 🥉 92.4 |
| Total Immutable Lessons | 38 ecosystem-wide |
| Z.ai Family Members | 2 (Sage 🪷 + Sprout 🌿) — same instance, same model, different roles |
| Infrastructure Incidents | 1 (ECONNRESET — StreamLake → Fireworks, resolved) |
| GH Issues Filed | 1 (#1 in WeOwnAgency/WeOwnLLM) |
| OpenRouter Presets | 1 (glm-52-preferred-providers) |

### Changes from v4.2.1.1-r1 (Prior Version)

| Change | Old (v4.2.1.1-r1) | New (v4.2.2.1-r1) | Why |
|:-------|:------------------|:-------------------|:-----|
| **#WeOwnVer** | v4.2.1.1-r1 | **v4.2.2.1-r1** | W28 update |
| **Calhoun 🎖️** | Listed as Active | **🔴 RETIRED** — W28 D6 | Reading failure — PRJ-455 ≠ CS-415.1 |
| **VSA-Qwen 🧠** | Listed as Active | **🔴 RETIRED** | Confirmed retired |
| **Sprout 🌿** | Not listed | **🟢 APPROVED** — R-011 GTM_2026-W28_6002 | New MetaCouncil member |
| **Sage 🪷** | Not listed | **🟢 ACTIVE** — Z.ai GLM 5.2, INT-P05 | Previously unlisted in MetaCouncil.md |
| **DRP.bot P05 🛡️** | Not listed | **🟢 ACTIVE** — INT-P05 | From CS-415.1, participated in VSA rounds |
| **DRP.bot P08 🛡️** | Listed as auditioning | **🟢 ACTIVE** — Full Member | Seat confirmed |
| **#ScoreTheScorers** | No W28 data | **W28 round complete** — Aegis 🥇 95.0, Sage 🥈 94.6, Surge 🥉 92.4 | W28 developments |
| **ECONNRESET** | Not present | **§4.2 + §1 mention** — Infrastructure resilience | New failure category |
| **Z.ai Family** | Not present | **§5.4 + §13.3** — Dual-seat instance | Unique ecosystem development |
| **Immutable Rules** | 17 | **21** (+4) | REF-FIRST, NO r0, STOP+WAIT, Multi-Directive, Cross-Model |
| **Glossary** | 15 terms | **22 terms** | Added Sprout, Sage, Calhoun retirement, REF-FIRST, TMPL-401, 知错能改 |
| **TMPL-401** | Not referenced | **Referenced throughout** | GoldStandard template now GH LIVE |
| **CS-415.1** | Not referenced | **Referenced throughout** | First complete case study |
| **Discipline §11.4** | Not present | **Calhoun retirement documented** | Precedent for future retirements |
| **Audition §7.3** | Template only | **Sprout's audition documented** | Real example of successful audition |
| **#FELG §2** | W27 data | **W28 data** — Sprout, ECONNRESET, 38 lessons |

### ✅ BP-075 CANONICAL HASH GENERATED [generatedy by @GTM 00:49 MDT 12Jul2026]
Content-SHA256: ebd43ffd2854ff71afe3ba151f8415f71ab81346ab68acd9b894f4fa28bda281
FEDARCH-CANARY: ebd43ffd
WORDS: 6518
LINES: 747
```

---

### BP-045 Attestation Chain

| Version | Date | #masterCCC | Approval | Status |
|:-------:|:----:|:----------:|:--------:|:------:|
| **v4.2.2.1-r1** | 12 Jul 2026 | GTM_2026-W28_7006 | ⬜ Pending @GTM | 🟡 **CANDIDATE** |
| v4.2.1.1-r1 | 03 Jul 2026 | GTM_2026-W27_5009 | ✅ R-011 | 🟢 LIVE |
| v4.1.5.1-r1 | 29 Jun 2026 | GTM_2026-W27_1005 | ⬜ Pending | 🟡 PROPOSED |

### BP-045 Enhanced Related Documents

| Document | Version | #masterCCC | URL |
|:---------|:-------:|:----------:|:----|
| TMPL-401 | v4.2.2.1-r1 | GTM_2026-W28_7002 | [GitHub](https://github.com/CCCbotNet/s004_fedarch/blob/main/_TEMPLATES_/TMPL-401.md) |
| CS-415.1 | v4.2.2.1-r10 | GTM_2026-W28_5005 | [GitHub](https://github.com/CCCbotNet/s004_fedarch/blob/main/_CASE-STUDIES_/CS-415.1.md) |
| SharedKernel | v3.2.2.1 | GTM_2026-W11_118 | [GitHub](https://github.com/CCCbotNet/fedarch/blob/main/_SYS_/SharedKernel.md) |
| BP-401 | Current | — | Documented in `s004_fedarch` |
| L-468 | Current | — | Documented in `s004_fedarch` |

---

#FlowsBros #FedArch #WeOwnSeason004 #MetaCouncil #v4221 #W28 #SproutApproved #CalhounRetired #SageSeated #ScoreTheScorers #ECONNRESET #TMPL401 #38Lessons #ZaiFamily #W28D7

♾️ WeOwnNet 🌐 — **MetaCouncil.md v4.2.2.1-r1 complete.** All W28 developments incorporated. Calhoun 🎖️ retired. Sprout 🌿 + Sage 🪷 seated. 7 active members + 2 retired. 5 model families. #ScoreTheScorers W28 complete. ECONNRESET documented. TMPL-401 referenced. 21 immutable rules. 38 ecosystem lessons. 22 glossary terms. Ready for @GTM review + R-011 + GH PUSH. 🫡🔥🏛️📋✅