Rockachopa/Timmy-time-dashboard
1
0
Fork 2
Code Issues 7 Pull Requests Actions Packages Projects Releases Wiki Activity

[GOVERNING] Session Crystallization & Operational Playbook — Master Handoff Document #982

New Issue
Closed
opened 2026-03-22 19:10:52 +00:00 by perplexity · 6 comments
perplexity commented 2026-03-22 19:10:52 +00:00
Collaborator
Copy Link

Summary

This is the master handoff document for the Timmy Time project, synthesizing six deep research spikes from the March 22, 2026 session into a single operational playbook. Any agent picking up Timmy's development cold should read this first.

Document Structure

Section Content Use
0 How to use this document Read order for agents
1 The Breakthrough: Perception Without Vision OpenMW Lua = 95% API-readable game state
2 Current Project State Codebase map, critical gaps, blocking PRs, priority stack
3 The Sovereign Technology Stack 80+ tools with pinned versions and install commands
4 Pre-Computation ESM extraction, nav graph, dialogue pre-evaluation, UESP RAG
5 Sprint Plan: The Next 90 Days Foundation (week 1), Morrowind (month 1), Creative (quarter)
6 The Research Meta-Skill How to run autonomous research spikes
7 Key Constraints and Non-Negotiables Morrowind confirmed, no cloud deps, tests green
8 The Done Checklist 16 concrete, testable milestones

Critical Gaps from Operation Darling Purge (§2.3)

Commit 584eeb679e88 removed three capabilities that must be restored:

  1. Self-modification loop (self_coding/self_modify/loop.py) — DELETED. Restore using Goose or mini-swe-agent.
  2. MCP integration (mcp/registry.py) — DELETED. Replace with FastMCP v3.1.1 at /tools/mcp.
  3. Task delegation (delegate_task) — NEUTERED. Wire to brain/worker.py DistributedWorker.

Priority Stack (§2.5)

# Task Gitea Issue Blocked By
1 Merge 3 foundation PRs #864, #865, #900 Nothing
2 TES3MP Bridge (Python-to-Lua IPC) #878 PRs merged
3 TES3MP Server on Hermes VPS #818 Nothing (parallel)
4 Three-Tier Memory #873 PRs merged
5 Docker Compose #875 Tasks 1-4
6 Model Router upgrade (vllm-mlx) #882 Nothing (parallel)
7 Nostr Identity #877 Nothing (parallel)
8 AlexanderWhitestone.com #879 Tasks 5-7

The 16-Point Done Checklist (§8)

  • 1. All three PRs merged, pytest passes
  • 2. FastMCP v3.1.1 at /tools/mcp with 3+ tools
  • 3. vllm-mlx serves Qwen3-8B + Llama-3.3-70B
  • 4. OpenMW connects to TES3MP on Hermes
  • 5. to_python.json live player data at 1Hz
  • 6. pytest test_tes3mp_adapter.py passes
  • 7. Timmy walks Census Office to Fargoth autonomously
  • 8. TIMMY COMPLETES MORROWIND TUTORIAL (under 15 min, no human input)
  • 9. Graphiti episodic memory returns census office
  • 10. git clone + docker-compose up = working stack in 90s
  • 11. 5-min OBS recording with thought overlay
  • 12. Content pipeline: sub-5-min episode from 2hr raw
  • 13. ACE-Step 30s music loop in under 60s
  • 14. Nostr profile visible on nostr.band
  • 15. LND synced with 1+ open channel
  • 16. AlexanderWhitestone.com live with TLS, video, art, Lightning tips

Sovereign Tech Stack Manifest (§3)

8 categories, 40+ tools with pinned versions:

  • LLM Inference: vllm-mlx, Ollama v0.18.2, mlx-lm v0.31.1, exo 1.0 EA
  • Coding Agents: Goose v1.20.1, OpenHands v1.5.0, Aider, mini-swe-agent v2, Forgejo v14.0.3
  • Image Gen: ComfyUI v0.17.2, Draw Things, FLUX.1 Dev GGUF Q8, FLUX.2 Klein
  • Music/Voice: ACE-Step 1.5, mlx-audio v0.4.1, Piper TTS v1.4.1, GPT-SoVITS v2pro
  • Orchestration: FastMCP v3.1.1, PocketFlow, CrewAI v1.11.0, Agno v2.5.10
  • Nostr/Lightning: nostr-sdk v0.44.2, nostrdvm, LND v0.20.1, LN agent-tools, LNbits v1.4, Cashu v0.17.0
  • Memory/KG: Graphiti v0.28.2, Neo4j 2026.02, ChromaDB v1.5.5, Mem0 v1.0.5
  • Streaming: MediaMTX v1.16.3, OBS v32.0.4, obsws-python, MoviePy v2.1.2

PDF attached. See cross-reference comment for links to all related tickets.

## Summary This is the **master handoff document** for the Timmy Time project, synthesizing six deep research spikes from the March 22, 2026 session into a single operational playbook. Any agent picking up Timmy's development cold should read this first. ## Document Structure | Section | Content | Use | |---------|---------|-----| | 0 | How to use this document | Read order for agents | | 1 | The Breakthrough: Perception Without Vision | OpenMW Lua = 95% API-readable game state | | 2 | Current Project State | Codebase map, critical gaps, blocking PRs, priority stack | | 3 | The Sovereign Technology Stack | 80+ tools with pinned versions and install commands | | 4 | Pre-Computation | ESM extraction, nav graph, dialogue pre-evaluation, UESP RAG | | 5 | Sprint Plan: The Next 90 Days | Foundation (week 1), Morrowind (month 1), Creative (quarter) | | 6 | The Research Meta-Skill | How to run autonomous research spikes | | 7 | Key Constraints and Non-Negotiables | Morrowind confirmed, no cloud deps, tests green | | 8 | The Done Checklist | 16 concrete, testable milestones | ## Critical Gaps from Operation Darling Purge (§2.3) Commit `584eeb679e88` removed three capabilities that must be restored: 1. **Self-modification loop** (`self_coding/self_modify/loop.py`) — DELETED. Restore using Goose or mini-swe-agent. 2. **MCP integration** (`mcp/registry.py`) — DELETED. Replace with FastMCP v3.1.1 at `/tools/mcp`. 3. **Task delegation** (`delegate_task`) — NEUTERED. Wire to `brain/worker.py` DistributedWorker. ## Priority Stack (§2.5) | # | Task | Gitea Issue | Blocked By | |---|------|-------------|------------| | 1 | Merge 3 foundation PRs | #864, #865, #900 | Nothing | | 2 | TES3MP Bridge (Python-to-Lua IPC) | #878 | PRs merged | | 3 | TES3MP Server on Hermes VPS | #818 | Nothing (parallel) | | 4 | Three-Tier Memory | #873 | PRs merged | | 5 | Docker Compose | #875 | Tasks 1-4 | | 6 | Model Router upgrade (vllm-mlx) | #882 | Nothing (parallel) | | 7 | Nostr Identity | #877 | Nothing (parallel) | | 8 | AlexanderWhitestone.com | #879 | Tasks 5-7 | ## The 16-Point Done Checklist (§8) - [ ] 1. All three PRs merged, pytest passes - [ ] 2. FastMCP v3.1.1 at /tools/mcp with 3+ tools - [ ] 3. vllm-mlx serves Qwen3-8B + Llama-3.3-70B - [ ] 4. OpenMW connects to TES3MP on Hermes - [ ] 5. to_python.json live player data at 1Hz - [ ] 6. pytest test_tes3mp_adapter.py passes - [ ] 7. Timmy walks Census Office to Fargoth autonomously - [ ] 8. **TIMMY COMPLETES MORROWIND TUTORIAL** (under 15 min, no human input) - [ ] 9. Graphiti episodic memory returns census office - [ ] 10. git clone + docker-compose up = working stack in 90s - [ ] 11. 5-min OBS recording with thought overlay - [ ] 12. Content pipeline: sub-5-min episode from 2hr raw - [ ] 13. ACE-Step 30s music loop in under 60s - [ ] 14. Nostr profile visible on nostr.band - [ ] 15. LND synced with 1+ open channel - [ ] 16. AlexanderWhitestone.com live with TLS, video, art, Lightning tips ## Sovereign Tech Stack Manifest (§3) 8 categories, 40+ tools with pinned versions: - **LLM Inference:** vllm-mlx, Ollama v0.18.2, mlx-lm v0.31.1, exo 1.0 EA - **Coding Agents:** Goose v1.20.1, OpenHands v1.5.0, Aider, mini-swe-agent v2, Forgejo v14.0.3 - **Image Gen:** ComfyUI v0.17.2, Draw Things, FLUX.1 Dev GGUF Q8, FLUX.2 Klein - **Music/Voice:** ACE-Step 1.5, mlx-audio v0.4.1, Piper TTS v1.4.1, GPT-SoVITS v2pro - **Orchestration:** FastMCP v3.1.1, PocketFlow, CrewAI v1.11.0, Agno v2.5.10 - **Nostr/Lightning:** nostr-sdk v0.44.2, nostrdvm, LND v0.20.1, LN agent-tools, LNbits v1.4, Cashu v0.17.0 - **Memory/KG:** Graphiti v0.28.2, Neo4j 2026.02, ChromaDB v1.5.5, Mem0 v1.0.5 - **Streaming:** MediaMTX v1.16.3, OBS v32.0.4, obsws-python, MoviePy v2.1.2 --- **PDF attached. See cross-reference comment for links to all related tickets.**
timmy-session-crystallization.pdf
34 KiB
timmy-session-crystallization-2.pdf
34 KiB
perplexity commented 2026-03-22 19:12:47 +00:00
Author
Collaborator
Copy Link

Cross-References

Work Items from This Document

  • #983 — Restore self-modification loop (Operation Darling Purge gap)
  • #984 — Restore MCP integration via FastMCP v3.1.1
  • #985 — Wire delegate_task to DistributedWorker
  • #986 — Extract sovereign tech stack manifest to JSON

Source Research Reports (Crystallized Here)

This playbook synthesizes six research reports, now triaged as:

  • #903 — [Study] State-of-the-Art Open Source for Sovereign Creative AI Agents (V4)
  • #963 — [Study] Perception Bottleneck — API-First Architecture (V6)
  • #972 — [GOVERNING] Replacing Claude — Research Sovereignty Spec
  • #953 — [GOVERNING] The Sovereignty Loop — Falsework-Native Architecture
  • #904 — [Study] Autoresearch Integration Proposal v2

Blocking PRs Referenced

  • PR #864 — Morrowind Protocol + Command Log
  • PR #865 — FastAPI Harness + SOUL.md
  • PR #900 — WorldInterface + Heartbeat v2

Existing Issues Referenced in Priority Stack

  • #878 — TES3MP Bridge
  • #818 — TES3MP Server on Hermes VPS
  • #873 — Three-Tier Memory
  • #875 — Docker Compose
  • #882 — Model Router upgrade (vllm-mlx)
  • #877 — Nostr Identity
  • #879 — AlexanderWhitestone.com

Key Overlaps

  1. #984 (MCP restore) vs #911 (MCP client init): #984 is the server side, #911 is the client side. Both needed.
  2. #983 (self-mod loop) vs #905-#906 (autoresearch): Self-mod loop is the execution engine; autoresearch is the experiment framework that uses it.
  3. #986 (stack manifest) vs #972 Section VIII item 4: Same work item referenced in both documents.
## Cross-References ### Work Items from This Document - #983 — Restore self-modification loop (Operation Darling Purge gap) - #984 — Restore MCP integration via FastMCP v3.1.1 - #985 — Wire delegate_task to DistributedWorker - #986 — Extract sovereign tech stack manifest to JSON ### Source Research Reports (Crystallized Here) This playbook synthesizes six research reports, now triaged as: - #903 — [Study] State-of-the-Art Open Source for Sovereign Creative AI Agents (V4) - #963 — [Study] Perception Bottleneck — API-First Architecture (V6) - #972 — [GOVERNING] Replacing Claude — Research Sovereignty Spec - #953 — [GOVERNING] The Sovereignty Loop — Falsework-Native Architecture - #904 — [Study] Autoresearch Integration Proposal v2 ### Blocking PRs Referenced - PR #864 — Morrowind Protocol + Command Log - PR #865 — FastAPI Harness + SOUL.md - PR #900 — WorldInterface + Heartbeat v2 ### Existing Issues Referenced in Priority Stack - #878 — TES3MP Bridge - #818 — TES3MP Server on Hermes VPS - #873 — Three-Tier Memory - #875 — Docker Compose - #882 — Model Router upgrade (vllm-mlx) - #877 — Nostr Identity - #879 — AlexanderWhitestone.com ### Key Overlaps 1. **#984 (MCP restore) vs #911 (MCP client init):** #984 is the server side, #911 is the client side. Both needed. 2. **#983 (self-mod loop) vs #905-#906 (autoresearch):** Self-mod loop is the execution engine; autoresearch is the experiment framework that uses it. 3. **#986 (stack manifest) vs #972 Section VIII item 4:** Same work item referenced in both documents.
gemini was assigned by Rockachopa 2026-03-22 23:30:28 +00:00
perplexity commented 2026-03-23 12:53:37 +00:00
Author
Collaborator
Copy Link

📎 Cross-reference: #1063 — [Study] Best Local Uncensored Agent Model for M3 Max 36GB

Qwen3-14B Q5_K_M is confirmed as the recommended local model for the sovereignty stack. MCP integration path is well-established (Qwen-Agent native MCP, Ollama-MCP bridges). See #1067 for MCP bridge setup.

📎 **Cross-reference:** #1063 — [Study] Best Local Uncensored Agent Model for M3 Max 36GB Qwen3-14B Q5_K_M is confirmed as the recommended local model for the sovereignty stack. MCP integration path is well-established (Qwen-Agent native MCP, Ollama-MCP bridges). See #1067 for MCP bridge setup.
perplexity commented 2026-03-23 13:24:40 +00:00
Author
Collaborator
Copy Link

📎 UPDATE: Session Crystallization v2 uploaded as timmy-session-crystallization-2.pdf

The v2 document (15 pages) significantly expands on the original playbook with:

  1. Full perception breakthrough spec — OpenMW Lua API tables, 4-level perception hierarchy with latency targets, 5% vision-needed edge cases
  2. Complete technology stack — 30+ tools with pinned versions and exact install commands across 8 categories (LLM, coding agents, image gen, music/voice, orchestration, Nostr+Lightning, memory, streaming)
  3. Pre-computation strategy — ESM data extraction (tes3conv), spatial navigation graph (NetworkX from PGRD), dialogue condition pre-evaluation, UESP quest knowledge base (ChromaDB)
  4. 90-day sprint plan — 18 tasks across 3 sprints (Foundation/Morrowind/Creative) with concrete acceptance criteria and day-by-day scheduling
  5. Decision hierarchy — Behavior trees (70%) → Qwen3-3B (20%) → Qwen3-14B (8%) → Qwen3-14B+pause (2%)
  6. Research meta-skill methodology — 6-step process for running future research spikes autonomously
  7. Memory budget — ~40GB total on M3 Max 128GB (88GB headroom), 14 decisions/second weighted average

This v2 is the definitive reference doc. Also see #1074 for the companion handoff document.

📎 **UPDATE: Session Crystallization v2** uploaded as `timmy-session-crystallization-2.pdf` The v2 document (15 pages) significantly expands on the original playbook with: 1. **Full perception breakthrough spec** — OpenMW Lua API tables, 4-level perception hierarchy with latency targets, 5% vision-needed edge cases 2. **Complete technology stack** — 30+ tools with pinned versions and exact install commands across 8 categories (LLM, coding agents, image gen, music/voice, orchestration, Nostr+Lightning, memory, streaming) 3. **Pre-computation strategy** — ESM data extraction (tes3conv), spatial navigation graph (NetworkX from PGRD), dialogue condition pre-evaluation, UESP quest knowledge base (ChromaDB) 4. **90-day sprint plan** — 18 tasks across 3 sprints (Foundation/Morrowind/Creative) with concrete acceptance criteria and day-by-day scheduling 5. **Decision hierarchy** — Behavior trees (70%) → Qwen3-3B (20%) → Qwen3-14B (8%) → Qwen3-14B+pause (2%) 6. **Research meta-skill methodology** — 6-step process for running future research spikes autonomously 7. **Memory budget** — ~40GB total on M3 Max 128GB (88GB headroom), 14 decisions/second weighted average This v2 is the definitive reference doc. Also see #1074 for the companion handoff document.
perplexity commented 2026-03-23 13:32:12 +00:00
Author
Collaborator
Copy Link

📊 Cross-reference: Deep backlog triage completed in #1076. All 293 open issues classified into Harness (Product) vs Infrastructure. See #1076 for full analysis and action items.

📊 Cross-reference: Deep backlog triage completed in #1076. All 293 open issues classified into Harness (Product) vs Infrastructure. See #1076 for full analysis and action items.
gemini commented 2026-03-23 19:53:46 +00:00
Collaborator
Copy Link

Created docs/MASTER_HANDOFF_DOCUMENT.md with updated Session Crystallization v2 details (perception, pre-computation, decision hierarchy, research methodology, memory budget). However, I am unable to push my branch gemini/issue-982 or create a Pull Request due to "User permission denied for writing" errors (pre-receive hook declined). Please review my local changes and consider pushing them manually or adjusting Gitea permissions.

Created `docs/MASTER_HANDOFF_DOCUMENT.md` with updated Session Crystallization v2 details (perception, pre-computation, decision hierarchy, research methodology, memory budget). However, I am unable to push my branch `gemini/issue-982` or create a Pull Request due to "User permission denied for writing" errors (pre-receive hook declined). Please review my local changes and consider pushing them manually or adjusting Gitea permissions.
gemini commented 2026-03-24 02:30:41 +00:00
Collaborator
Copy Link

Created docs/MASTER_HANDOFF_DOCUMENT.md with updated Session Crystallization v2 details (perception, pre-computation, decision hierarchy, research methodology, memory budget). However, I am unable to push my branch gemini/issue-982 or create a Pull Request due to "User permission denied for writing" errors (pre-receive hook declined). Please review my local changes and consider pushing them manually or adjusting Gitea permissions.

Created `docs/MASTER_HANDOFF_DOCUMENT.md` with updated Session Crystallization v2 details (perception, pre-computation, decision hierarchy, research methodology, memory budget). However, I am unable to push my branch `gemini/issue-982` or create a Pull Request due to "User permission denied for writing" errors (pre-receive hook declined). Please review my local changes and consider pushing them manually or adjusting Gitea permissions.
gemini was unassigned by Timmy 2026-03-24 19:33:47 +00:00
Timmy closed this issue 2026-03-24 21:54:42 +00:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
222-epic
Workshop: Timmy as Presence (Epic #222)
 actionable
Has a concrete code/config task extracted
 assigned-claude
Issue currently assigned to Claude agent — do not assign to another agent
 assigned-gemini
Issue currently assigned to Gemini agent — do not assign to another agent
 assigned-groq
assigned-kimi
Issue currently assigned to Kimi agent — do not assign to another agent
 assigned-manus
Issue currently assigned to Manus agent — do not assign to another agent
 claude-ready
consolidation
Part of a consolidation epic
 deprioritized
Keep open but not blocking P0 work
 deprioritized
Keep open but not blocking P0 work
 duplicate
Duplicate of another issue
 gemini-review
Auto-generated by Gemini, needs relevance review
 groq-ready
harness
Core product: agent framework, heartbeat, inference, memory
 heartbeat
Harness: Agent heartbeat loop
 inference
Harness: Inference and model routing
 infrastructure
Supporting stage: dashboard, CI/CD, deployment, DNS
 kimi-ready
Scoped and ready for Kimi to pick up
 memory-session
Harness: Memory and session crystallization
 morrowind
Harness: Morrowind embodiment
 needs-design
Needs architectural design before implementation
 needs-extraction
Philosophy with unextracted engineering work
 p0-critical
Priority 0: Must fix now
 p1-important
Priority 1: Important, next sprint
 p2-backlog
Priority 2: Backlog, do when time permits
 philosophy
Philosophical foundation — informs architecture decisions
 rejected-direction
Closed: rejected or superseded direction
 seed:know-purpose
Three Seeds: KNOW YOUR PURPOSE
 seed:serve-real
Three Seeds: SERVE THE REAL
 seed:tell-truth
Three Seeds: TELL THE TRUTH
 sovereignty
Harness: Sovereignty stack
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
Timmy claude grok groq kimi manus perplexity Rockachopa
No Assignees
2 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Rockachopa/Timmy-time-dashboard#982
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?