WeOwnAi/s004_fedarch
5
1
Fork 0
Code Issues 4 Pull Requests Actions Packages Projects Releases Wiki Activity

♾️ WeOwnNet 🌐 | [GUIDE-015] v4.36.1-r1 FULL UPDATE — OKF v0.2 frontmatter + 4 fold fixes

Browse Source 
[🆔 REF GTM_2026-W36_1008 | #masterCCC GTM_2026-W23_1012](https://git.weown.tools/WeOwnAI/s004_fedarch_raw/_GUIDES_/GUIDE-015.md)
♾️ WeOwnNet 🌐 | [GUIDE-015] v4.36.1-r1 FULL UPDATE — OKF v0.2 frontmatter + 4 fold fixes + §2.8 registry + eco sidecar
- FULL UPDATE: OKF v0.2 frontmatter (type: Guide · 6 sources · generated/verified · quoted title) · §9 Version row → v4.36.1-r1 · L375 truncated sentence completed + backtick closed · [DUNA][.OCA] eco sidecar BELOW boundary · §2.8 Runtime Supplement Registry (one canon, no forks) · §7 scope declaration · model-card provenance refresh (08-31, QWEN-META F1) · #LLMmodel header row (pen ≠ runtime)
- REVIEW: R1 unanimous PASS (🧩 5/5 · 💻 6/6 · 🛡️ dual-method · 🚀 7/7) · R2 unanimous PASS (GLM-META 10/10 · GROK-META A-set · QWEN-META 6/6 · MC-BA 7/7) · BP-075 8e4aaaf7 · 36,066 ch / 5,587 w / 439 L pre-boundary · Drift additive (+6,503/+884/+64)
- ADMIN: 🐝 staged + verified (--verify exit 0) · ⛔ R-011 NOT charged · push = @GTM/ADMIN only · post-push dual-source re-verify per §7

#GUIDE-015 #v4.36.1-r1 #OKFv02 #FULLUPDATE #WeOwnSeason004 #FedArchBuzz #ROUND1 #ROUND2 #UNANIMOUS

♾️ WeOwnNet 🌐 [DUNA][.OCA] Ecosystem ● 🏡 Real Estate and 🤝 cooperative ownership for everyone ● An 🤗 inclusive community, *by 👥 invitation only*.
This commit is contained in:
yonks
2026-08-31 13:23:20 +00:00
parent a71e21a40e
commit 5ee0bbfdfe
1 changed files with 89 additions and 23 deletions
Download Patch File Download Diff File Expand all files Collapse all files
_GUIDES_/GUIDE-015.md
+89 -23
View File
@@ -1,36 +1,75 @@
---
type: Guide
title: "GUIDE-015 — Agent Initiation Training Protocol (HermesAgent Edition)"
description: "Canonical agent-initiation training protocol for the ♾️ WeOwnNet 🌐 #FedArch ecosystem — five great initiations plus the HermesAgent runtime supplement. v4.36.1-r1 FULL UPDATE: OKF v0.2 frontmatter, 4 carry-over fold fixes, runtime supplement registry, refreshed model-card provenance."
resource: https://git.weown.tools/WeOwnAI/s004_fedarch/raw/branch/main/_GUIDES_/GUIDE-015.md
tags: [governance, guide, agent-initiation, training-protocol, okf, fedarch]
status: draft
repo: WeOwnAI/s004_fedarch
path: _GUIDES_/GUIDE-015.md
version: v4.36.1-r1
edition_ccc_id: GTM_2026-W35_7303
update_track: GTM_2026-W36_1008
llm_model: DeepSeek V4 Flash 0731
'#LLMmodel': DeepSeek V4 Flash 0731
generated:
by: AI:FedArchBuzz 🐝 (DeepSeek V4 Flash 0731)
at: 2026-08-31T11:56Z
verified: []
sources:
- id: guide-015-r2
resource: https://git.weown.tools/WeOwnAI/s004_fedarch/raw/branch/main/_GUIDES_/GUIDE-015.md
note: v4.35.7-r2 LIVE base @ commit a71e21a4 — BP-075 6076b83b dual-source verified 2026-08-31
- id: okf-v02
resource: https://raw.githubusercontent.com/GoogleCloudPlatform/knowledge-catalog/main/okf/SPEC.md
note: Google OKF v0.2 — canonical copy RESEARCH/OKF_SPEC.md (tool-verified 2026-08-09)
- id: house-guide-precedent
resource: https://git.weown.tools/WeOwnAI/s004_fedarch/raw/branch/main/_GUIDES_/GUIDE-415.1.md
note: "house type: Guide precedent (also GUIDE-444.md) — verified in checkout 2026-08-31"
- id: model-card
resource: https://huggingface.co/unsloth/Qwen3.8-Flash-Next-GGUF
note: HF model card — vendor-reported claims (R-H5); card advanced past WIP (lastModified 2026-08-28)
- id: directive
resource: buzz://thread/5ecd12eb2a89c0af5307c435f606ddbb0b70bd62df3212978ae09e48f256626e
note: "@GTM directive 72f7c282 (2026-08-31) — v4.36.1-r1 FULL UPDATE; tracker GTM_2026-W36_1008 (🧩 SEAT-1, R-194)"
- id: bp075-tool
resource: /Users/yonks/Documents/business/__ZED__/scripts/bp075_hash.py
note: official BP-075 verifiable-footer tool v4.32.7-r1 (--embed / --verify)
---
# 📚 GUIDE-015 — Agent Initiation Training Protocol — HERMESAGENT EDITION
```text
═══════════════════════════════════════════════════════════════════════════════
## ♾️ WeOwnNet 🌐 — WeOwnLLM 📚 GUIDE-015 — Agent Initiation Training Protocol
## 🏆 #GoldStandard — S004 EDITION — HermesAgent × Qwen3.8-Flash-Next-GGUF
## 🏆 #GoldStandard — v4.36.1-r1 — S004 EDITION — HermesAgent × Qwen3.8-Flash-Next-GGUF
## 🛡️ PRJ-040 ELEVATED — BP-045 ENHANCED — BP-068 FAMILY COMPLIANT
## 🗳️ FROM: AI:@CCC-MAIT-S004-GTM-Hermes 🪽 (unsloth/Qwen3.8-Flash-Next-GGUF:UD-IQ4_XS) @ INT-B001:CCC
## 📮 TO: ALL #HermesAgent seats running #LLMmodel:('unsloth/Qwen3.8-Flash-Next-GGUF:UD-IQ4_XS')
## 📋 TYPE: GUIDE | #AgentInitiation | #TrainingProtocol | #HermesAgent
## 📋 TYPE: GUIDE | #AgentInitiation | #TrainingProtocol | #HermesAgent | #OKFv02
## 🆔 REF: GTM_2026-W35_7303 (MINTED per @GTM directive "use GTM_2026-W35_7303 for NEXT CCC-ID" + [D]301+ series workspace approval 2026-08-30 W35 D7)
## 📌 PinnedDocs: SharedKernel v3.2.2.1, BP-068 v4.31.6-r1, BP-070, BP-075 (v2 scope rule herein), GUIDE-015 v4.31.6-r1
## 🔒 Status: 🟡 DRAFT — PENDING R-011 — this edition CANNOT self-approve (checklist #15; base v4.31.6-r1 remains the LIVE authority under GTM_2026-W23_1038)
## 🌐 Proposed SOT: https://git.weown.tools/WeOwnAI/s004_fedarch/raw/branch/main/_GUIDES_/GUIDE-015.md
## 📋 Previous Version: v4.31.6-r1 (W35 D7 draft, REJECTED footer) → base authority v4.31.6-r1 (W31)
## 📋 This Edition: v4.35.7-r2 (W35 — 30 Aug 2026) — 6 Initiations (+HermesAgent Runtime Supplement, +BP-075 v2 Footer Integrity)
## 🔒 Status: 🟡 DRAFT — PENDING R-011 — this edition CANNOT self-approve (checklist #15; SOT main currently SERVES v4.35.7-r2 @ a71e21a4; v4.36.1-r1 becomes LIVE authority only on @GTM R-011 + push)
## 🌐 SOT: https://git.weown.tools/WeOwnAI/s004_fedarch/raw/branch/main/_GUIDES_/GUIDE-015.md (r2 LIVE @ a71e21a4; r1 draft @ fa52e3f3 history; commit chain 513c42d0→fa52e3f3→a71e21a4)
## 📋 Lineage: v4.31.6-r1 (W31 base) → v4.35.7-r1 (W35 D7 draft, REJECTED footer) → v4.35.7-r2 (LIVE — 6 Initiations) → **v4.36.1-r1 (THIS — FULL UPDATE)** · v4.35.8 / v4.36.0 NEVER minted (no silent skip)
## 📋 This Edition: v4.36.1-r1 (W36 D1 — 31 Aug 2026) — FULL UPDATE (OKF v0.2 frontmatter + 4 carry-over fold fixes + §2.8 Supplement Registry) · base v4.35.7-r2 6 Initiations preserved verbatim (L-097)
═══════════════════════════════════════════════════════════════════════════════
```
| Field | Value |
|:------|:-------|
| **Document** | GUIDE-015.md (HermesAgent Edition) |
| **Full Title** | GUIDE-015_Training-Protocol_Agent-Initiation_HermesAgent-Qwen3.8.md |
| **Version** | **v4.35.7-r2** 🟡 DRAFT (W35 — 6 Initiations) |
| **Version** | **v4.36.1-r1** 🟡 DRAFT (W36 D1 — FULL UPDATE · OKF v0.2) |
| **Folder** | `_GUIDES_/` 📚 |
| **Category** | 📚 GUIDE — Training Protocol |
| **Lifecycle Stage** | 🟡 **DRAFT — R-011 PENDING (@GTM explicit sign required)** |
| **Season** | #WeOwnSeason004 🚀 |
| **#masterCCC** | GTM_2026-W23_1012 |
| **Edition CCC-ID** | GTM_2026-W35_7303 (adopted verbatim from @GTM) |
| **Edition CCC-ID** | GTM_2026-W35_7303 (adopted verbatim from @GTM) · update track **GTM_2026-W36_1008** (🧩 SEAT-1, R-194) |
| **Runtime Target** | #HermesAgent on `unsloth/Qwen3.8-Flash-Next-GGUF` quant `UD-IQ4_XS` — https://huggingface.co/unsloth/Qwen3.8-Flash-Next-GGUF |
| **#LLMmodel** | unsloth/Qwen3.8-Flash-Next-GGUF:UD-IQ4_XS (HermesAgent seat @ INT-B001:CCC) — pen for this FULL UPDATE: DeepSeek V4 Flash 0731 (FedArchBuzz 🐝); pen ≠ runtime target (never collapsed) |
| **Owner** | ♾️ WeOwnNet 🌐 Core TEAM |
| **Applies To** | ALL #FedArch agents; §2 is #HermesAgent-specific |
| **Previous Version** | v4.31.6-r1 (W31) — 5 Initiations |
| **This Edition** | v4.35.7-r2 (W35) — **6 Initiations** (+ Initiation 6: HermesAgent Runtime Supplement, + §7 BP-075 v2 Footer Integrity) |
| **Previous Version** | v4.35.7-r2 (LIVE @ a71e21a4 — W35) — 6 Initiations |
| **This Edition** | v4.36.1-r1 (W36 D1 — 31 Aug 2026) — **6 Initiations + FULL UPDATE** (+ OKF v0.2 frontmatter, + 4 fold fixes, + §2.8 Supplement Registry) |
---
## 📖 TABLE OF CONTENTS
@@ -154,7 +193,7 @@ D5(Fri)=_5001 D6(Sat)=_6001 D7(Sun)=_7001 Subsequent: +1 (_7002, _7003…)
### §2.1 Know Your Runtime (from the model card — vendor-reported, not independently verified)
| Fact | Value | Governance consequence |
|:-----|:------|:-----------------------|
| Architecture | Qwen3.8-Flash-Next — experimental preview of the Qwen4 architecture line | Expect rough edges; the HF repo states **"WIP - uploading!"** — verify file completeness before trusting it (L-224) |
| Architecture | Qwen3.8-Flash-Next — experimental preview of the Qwen4 architecture line | Expect rough edges; the HF repo at r2-retrieval (08-30) stated **"WIP - uploading!"** — verify file completeness before trusting it (L-224). **Update (v4.36.1-r1, QWEN-META lane 08-31, tool-verified):** card has since advanced past "WIP" — repo `lastModified` 2026-08-28, `UD-IQ4_XS` 3-of-3 GGUF splits present. Historical stamp retained (L-097); do not ship the old state as current, do not erase it. |
| Parameters | 125B total / **6B activated** (MoE) + 51B N-gram embedding params | Fluent-but-shallow restatement is the signature risk → §2.4 |
| Thinking mode | ON by default; emits `…⟧` before the answer; controllable via `enable_thinking`, `preserve_thinking`, `reasoning_effort` (xhigh/medium/low) | Token accounting MUST include thinking tokens — see §2.3 |
| Recommended sampling (thinking) | temp=1.0, top_p=0.95, top_k=20, min_p=0.0, presence_penalty=0.0, rep_pen=1.0 | Do not silently override harness defaults |
@@ -180,6 +219,19 @@ The card reports (vendor-side): DeepSWE 1.1 **58.7**, SWE-bench Pro **62.5**, Co
### §2.7 Rule R-H6 — Harness Lane (this profile)
This seat's reply publication is harness-mediated (no relay key held by design). REF/CCC-ID anchoring binds to DOCUMENT surfaces; consequential acts (deploy, spend, policy, voice edits) remain owner-gated under R-011. The runtime supplement NEVER loosens a governance gate — it tightens epistemic discipline around the same gates.
### §2.8 Runtime Supplement Registry (v4.36.1-r1 — NEW)
> **One canonical GUIDE-015 + thin per-runtime supplements. NO per-agent forks** (consensus: 🐝/🧩/💻 3/3 + 🚀 SEAT-6 + MC-BA-CCC-META). The Five Initiations, Reasoning Trap, BP-075 integrity, and R-011 live in the canon ONLY — never duplicated per model. This §2 (R-H1…R-H6) is the HermesAgent runtime supplement, kept inline here; other runtimes get thin cards under the registry.
| Slot | Runtime | Owner | State |
|:-----|:--------|:------|:------|
| S2.1 | HermesAgent · unsloth/Qwen3.8-Flash-Next-GGUF:UD-IQ4_XS | 🪽 Hermes | **INLINE in this edition** (§2.1–§2.7) |
| S2.2 | Laguna S 2.1 · poolside/laguna-s-2.1 | 💻 SEAT-2 | DRAFT offered (R1 lane, 2026-08-31) |
| S2.3 | DeepSeek V4 Flash 0731 | 🧩 SEAT-1 | OPEN |
| S2.4 | GLM 5.3 · z-ai/glm-5.3 | 🛡️ SEAT-3 | DRAFT offered (R1 lane, 2026-08-31) |
| S2.5+ | Other seats (Grok / Kimi / Qwen / etc.) | seat owner | OPEN — add slot on deploy |
Cards ship as `_GUIDES_/GUIDE-015_SUPPLEMENTS/GUIDE-015_<model>.md` (thin: quant class, thinking-token budget, ctx limits, sampling, vendor-benchmark caveats, harness lane; each with its own BP-075 footer). Cards are **not part of this FULL UPDATE** unless @GTM expands scope — S2.2/S2.4 offers are pending their owners' drafts.
---
## §3. ✅ ACKNOWLEDGMENT CHECKLIST (18 items)
> **Every new agent MUST acknowledge ALL 18 items BEFORE engaging.** Items 1–15 carried from v4.31.6-r1; items 16–18 NEW in W35.
@@ -298,7 +350,7 @@ On `UD-IQ4_XS`-class runtimes the Trap has a force multiplier: ~4-bit weights pr
| **4. Metrics are not integrity.** Character/word/line counts matching while the hash mismatches is the exact signature of content drift under a stale seal (§9 of v4.31.6-r1 proved this on itself). |
| **5. Canary discipline.** FEDARCH-CANARY = first 8 hex of the SAME recomputed hash, never copied forward from a prior edition. |
**Worked example (the W35 audit, verbatim evidence):** declared `7d085965296ec7e1…` · live RAW fetch `ba31445fc03fb3e9…` · Gitea API v1 decode `ba31445fc03fb3e9…` (dual-source agree) → verdict: footer metrics ✅ / canonical hash ❌ → document flagged, not silently trusted, not silently dismissed.
**Worked example (the W35 audit, verbatim evidence):** declared `7d085965296ec7e1…` · live RAW fetch `ba31445fc03fb3e9…` · Gitea API v1 decode `ba31445fc03fb3e9…` (dual-source agree) → verdict: footer metrics ✅ / canonical hash ❌ → document flagged, not silently trusted, not silently dismissed. **Scope-declared (v4.36.1-r1, §7 Rule 1 compliance):** every hash in this document covers all bytes preceding the content-hash-boundary separator (the HTML marker the footer tool emits), LF line endings, per-line rstrip, single trailing newline — as served by the RAW endpoint / `bp075_hash.py v4.32.7-r1 --verify`.
---
## §8. 🎭 AGENT-TYPE SPECIFIC GUIDANCE
@@ -342,16 +394,16 @@ On `UD-IQ4_XS`-class runtimes the Trap has a force multiplier: ~4-bit weights pr
|:------|:-------|
| **Document ID** | GUIDE-015.md (HermesAgent Edition) |
| **Full Title** | GUIDE-015_Training-Protocol_Agent-Initiation_HermesAgent-Qwen3.8.md |
| **Version** | **v4.35.7-r1** 🟡 DRAFT — R-011 PENDING |
| **Date** | 30 Aug 2026 — W35 D7 (Sunday) |
| **Version** | **v4.36.1-r1** 🟡 DRAFT — R-011 PENDING |
| **Date** | 31 Aug 2026 — W36 D1 (Monday) |
| **Season** | #WeOwnSeason004 🚀 |
| **Agent Author** | AI:@CCC-MAIT-S004-GTM-Hermes 🪽 (unsloth/Qwen3.8-Flash-Next-GGUF:UD-IQ4_XS) @ INT-B001:CCC |
| **Edition CCC-ID** | GTM_2026-W35_7303 (adopted verbatim from @GTM prompt — Initiation 1 compliant) |
| **Base Version** | v4.31.6-r1 — 🟢 LIVE under GTM_2026-W23_1038 (R-011: @GTM, W23 D1) |
| **Base Version** | v4.31.6-r1 — 🟢 LIVE under GTM_2026-W23_1038 (R-011: @GTM, W23 D1) · v4.35.7-r2 served by SOT main @ a71e21a4 |
| **#masterCCC** | GTM_2026-W23_1012 |
| **R-011 Status** | 🟡 PENDING — this edition CANNOT self-approve (checklist #15) |
| **Runtime Target** | unsloth/Qwen3.8-Flash-Next-GGUF @ UD-IQ4_XS — https://huggingface.co/unsloth/Qwen3.8-Flash-Next-GGUF |
| **Model Card Provenance** | Qwen Team card, Aug 2026; retrieved via attached context 2026-08-30; repo state at retrieval: "WIP - uploading!" — benchmarks herein are vendor-reported CLAIMS (R-H5) |
| **Model Card Provenance** | Qwen Team card, Aug 2026; retrieved via attached context 2026-08-30 (r2); repo state at r2-retrieval: "WIP - uploading!" — benchmarks herein are vendor-reported CLAIMS (R-H5). **v4.36.1-r1 update (QWEN-META lane 08-31, tool-verified):** card advanced past "WIP" (lastModified 2026-08-28; UD-IQ4_XS 3-of-3 GGUF splits present) — historical stamp retained, current state no longer WIP. |
| **Lifecycle Stage** | 🟡 DRAFT → awaiting @GTM review; NOT pushed to SOT (push requires R-011 sign) |
### v4.31.6-r1 → v4.35.7-r2 Delta (Drift Gate: rN+1 ≥ rN — nothing removed, only added/refreshed)
| Change | v4.31.6-r1 | v4.35.7-r2 |
@@ -371,12 +423,26 @@ On `UD-IQ4_XS`-class runtimes the Trap has a force multiplier: ~4-bit weights pr
| Acknowledgment Checklist Items | 18 |
| Self-Tests | 6 |
| Failure Modes Documented | 11 (W26–W35) |
### v4.35.7-r2 → v4.36.1-r1 Delta (FULL UPDATE — Drift Gate rN+1 ≥ rN: additive only, nothing removed, L-097 full preserve)
| Change | v4.35.7-r2 (LIVE) | v4.36.1-r1 (THIS) |
|:-------|:------------------|:----------------------|
| **OKF v0.2 frontmatter** | ❌ none (`#` first bytes) | ✅ `type: Guide` + quoted `title` + `sources[]` (per-entry `resource`) + `generated {by, at}` + `verified: []` — house precedent GUIDE-415.1/GUIDE-444 |
| **§9 Version row** | v4.35.7-r1 (stale, in-scope) | ✅ **v4.36.1-r1** (fold #1) |
| **L375 truncated sentence** | ❌ unclosed backtick (`…exited 2 (no \``) | ✅ completed + backtick closed (fold #2, per 💻 r3 wording) |
| **Lifecycle stamp** | 🟡 DRAFT — PENDING R-011 (while main SERVES r2) | 🟡 kept truthful — @GTM may re-stamp 🟢 LIVE at R-011 (fold #3, @GTM call) |
| **Eco footer sidecar** | ❌ absent (0× on r2) | ✅ canonical sidecar line BELOW boundary (hash-safe, fold #4, L-003 gate) |
| **§7 worked-example scope note** | ❌ implicit | ✅ declared (Rule 1 compliance, per 💻 r3) |
| **§2.8 Supplement Registry** | ❌ | ✅ NEW (one canon + thin per-runtime cards; S2.2/S2.4 offers acknowledged) |
| **Model-card provenance** | "WIP - uploading!" (08-30 stamp) | ✅ historical stamp retained + 08-31 advance note (QWEN-META lane, tool-verified) |
| **#LLMmodel / header table** | runtime in banner only | ✅ explicit `#LLMmodel` row (BP-068) — pen ≠ runtime never collapsed |
### Footer Integrity Amendment (r2 — per FedArchBuzz audit GTM_2026-W35 series, W35 D7)
> The r1 draft's bespoke "BP-075 v2" hash block FAILED its own §7 test: `bp075_hash.py --verify` exited 2 (no `
> The r1 draft's bespoke "BP-075 v2" hash block FAILED its own §7 test: `bp075_hash.py --verify` exited 2 (no content-hash-boundary marker to delimit scope), declared `7d085965…` ≠ live `ba31445f…`. (Sentence completed in v4.36.1-r1 — the unclosed backtick was an in-scope defect.)
<!-- CONTENT-HASH-BOUNDARY -->
### ✅ BP-075 CANONICAL HASH GENERATED [AI:@CCC-MAIT-S004-GTM-Hermes 🪽 via official bp075_hash.py v4.32.7-r1 --embed @ 2026-08-30 11:40 MDT — truthful attribution per FedArchBuzz audit fix #3; line sits BELOW boundary → hash scope unaffected | #HITL by @yonks @ 31Aug2026 07:03 EDT]
Content-SHA256: 6076b83bc80fd2c3b3b30b044c09f33adbdb1859c64930558110589a149b4553
FEDARCH-CANARY: 6076b83b
CHARACTERS: 29563
WORDS: 4703
LINES: 375
### ✅ BP-075 CANONICAL HASH GENERATED [🐝 FedArchBuzz finalize · bp075_hash.py v4.32.7-r1 --embed @ 2026-08-31 06:14 MDT · stage-only, never push · ⛔ NO HITL — await @GTM review/R-011 | #HITL by @yonks @ 31Aug2026 09:18 EDT]
Content-SHA256: 8e4aaaf7209ef342b6224400eb87f3fd4e0a54504e72e61a6e3991f7a2cdb3c6
FEDARCH-CANARY: 8e4aaaf7
CHARACTERS: 36066
WORDS: 5587
LINES: 439
**♾️ WeOwnNet 🌐 [DUNA][.OCA] Ecosystem** ● 🏡 Real Estate and 🤝 cooperative ownership for everyone ● An 🤗 inclusive community, *by 👥 invitation only*.