---
vault_clearance: EUCLID
  halo.created: auto
# HALO Protocol
> **H**onest. **A**ccessible. **L**iving. **O**urs.

## What This Is

Every entity in the lab — a cell culture, a dataset, a theory, a tool, a finding — gets a **Living Document** that follows a standardized structure but never stops growing. The document IS the entity's life story.

## The Living Document Format

Each Living Document has two layers:

### Layer 1: The Paper (public, growing)

| Section | What goes here | When it updates |
|---------|---------------|-----------------|
| **Abstract** | One paragraph. What this is, why it matters, what it showed. Pointers to lineage (parent experiments, related entities). | Every major finding |
| **Methods** | How it functions, behaves. Exact protocol, parameters, code paths. | When protocol changes |
| **Data** | Engines run, records, databases, raw observations. File locations, sizes, formats. | Every new data point |
| **Discussion** | Interpretation. Links to WORLDLINE sections where findings were analyzed. | After each analysis |
| **Conclusion** | What it means for the lab. What it means for the world. | When understanding deepens |
| **Next Steps** | Bounty board listings. New questions generated. | Continuously |

### Layer 2: The Containment (operational, security)

| Section | What goes here |
|---------|---------------|
| **Classification** | How sensitive is this? PUBLIC (can share), INTERNAL (lab only), RESTRICTED (key personnel), CLASSIFIED (encrypted) |
| **Containment** | What do we keep secret? What's the output format? What gets shared vs held? |
| **Addenda** | Append-only observations. Anomalies. Things that don't fit. "We noticed X but don't know what it means yet." |
| **Breach Protocol** | What happens if containment fails? If the data leaks? If someone scoops us? Contingency plan. |

## The Four Principles

### H — Honest
- State what we know AND what we don't
- Flag confidence levels (HIGH, MEDIUM, LOW, SPECULATIVE)
- Document failures and dead ends, not just successes
- If a dataset has caveats (P5-8 HUVECs, snRNA-seq missing mito), say so
- Never claim more than the data supports

### A — Accessible
- Any AI agent reading the project **BOOK** (and any registry subsection, e.g. §5) should understand the project in 30 seconds
- Any human collaborator should be able to run the pipeline in 3 commands
- File names describe their contents (no `data_final_v2_REAL.h5ad`)
- Living Documents link to each other through pointers, not assumptions
- The structure is self-documenting — the folder hierarchy IS the ontology

### L — Living
- Documents grow by APPENDING, not rewriting history
- Addenda are timestamped and signed (who observed this, when)
- The worldline is append-only — what happened stays in the record
- New findings update the Abstract and Conclusion but don't erase the old ones
- Dead ends get documented, not deleted — they're data too

### O — Ours
- The protocol belongs to the lab, not to any platform
- Works on any machine, any OS, any editor
- No vendor lock-in (GitHub, Notion, Obsidian are TOOLS, not dependencies)
- The encryption is semantic — the naming conventions ARE the code
- New collaborators earn access through contribution, not credentials

## File Structure Under HALO

```
Project/
  BOOK.md                # Bibliography + methods (+ fused registry subsection, e.g. §5 Combined Codex)
  QUICK_START.md         # 3-command onboarding
  README.md              # Public-facing overview
  WORLDLINE.md  # Append-only discovery log
  BOUNTY_BOARD.md        # Open questions and tasks
  HALO_PROTOCOL.md       # This file (living doc template: vault README § Living document template)

  methods/               # Code — the engines
    run_pipeline.py      # Main entry point
    *.py                 # Analysis modules

  data/
    experiments/         # Each experiment = one Living Document
      EXP01_.../
        LIVING_DOCUMENT.md
        DATA_LOCATION.txt
      EXP02_.../
        ...
    imaging/             # Microscopy — organized by condition + date
      YYYY-MM-DD_CellType_Condition/
        img_NNN.jpg
        vid_NNN.avi
    pathways/            # Gene set databases (.gmt)
    plots/               # Generated figures

  orthodox/              # Standard pipeline comparison (self-contained)
```

## How to Create a New Living Document

When you start a new experiment, dataset analysis, or theoretical exploration:

1. Create the folder: `data/experiments/EXPNN_descriptive_name/`
2. Start from the vault canonical template: [README § Living document template (HALO)](../README.md#living-document-template-halo), or copy a skeleton from [99_Archive/.../LIVING_DOCUMENT_TEMPLATE_symphony_variant.md](../99_Archive/root_reports/2026-03/unification_documentation_hardening/LIVING_DOCUMENT_TEMPLATE_symphony_variant.md) into `LIVING_DOCUMENT.md`
3. Fill in the Abstract (what are you doing and why)
4. Fill in Methods (exact protocol)
5. Set the Classification level
6. Add a pointer or row in **`BOOK.md`** (registry subsection where the project keeps one, e.g. §5)

Everything else grows organically as results come in.

## How to Read the Lab

1. **Start with `BOOK.md` / README** — find what exists (registry subsection if applicable)
2. **Read the Living Document** for any entity you care about
3. **Check WORLDLINE.md** for the analysis history
4. **Check BOUNTY_BOARD.md** for open questions
5. **Run the pipeline** if you want to verify anything

## The Worldline

The WORLDLINE.md file is the worldline — the append-only history of everything discovered. It never gets rewritten. New parts get added. Old parts stay. The worldline IS the lab's memory.

The Addenda in each Living Document are local worldlines — the history of that specific entity. Together, they form the complete record.

## Classification Tiers

| Tier | Symbol | Meaning | Sharing Rule | Effort to Protect |
|------|--------|---------|-------------|-------------------|
| **SAFE** | `[S]` | Show anyone. Public knowledge. | GitHub, Zenodo, papers | None |
| **EUCLID** | `[E]` | Uncertain response. Needs framing. | Publish with defensive context | Review before sharing |
| **KETER** | `[K]` | Danger to the lab if leaked. | Vault only. Core team. | Active containment |
| **APOLLYON** | `[A]` | Danger to world if leaked without defense. | Internal only. Defense published FIRST. | Maximum containment |
| **THAUMIEL** | `[T]` | Benefit to world. Should be shared. | Publish when WING criteria met. | Release when ready |

### Classification Flow

```
KETER (new finding)
  -> understand it -> EUCLID (theory with context)
  -> build defense -> THAUMIEL (cure/tool ready to share)
  -> attack surface -> APOLLYON (until defense published)
  -> defense published -> downgrade attack to EUCLID
  -> fully public -> SAFE
```

### Current Project Classifications

**SAFE:**
- Pipeline code (`methods/*.py`)
- Public dataset analyses (coupling hierarchy, cross-species)
- Project documentation (README, BOOK, QUICK_START)

**EUCLID:**
- Desync cascade theory
- Operator simplex theory
- Three-kingdom structure
- GBM operator victory state analysis
- Cross-species coupling hierarchy

**KETER:**
- All wet lab work, protocols, experimental designs
- iPSC coculture observations (`data/imaging/`)
- lncRNA checkpoint discovery (specific gene IDs)
- LIANA communication map (specific interactions)
- BAM file analyses
- Unpublished experimental results

**APOLLYON:**
- Chr19 UHRF1-DNMT1 TE insertion map (BloodyEchoes)
- Viral operator coupling analysis
- Synthetic TE insertion approach
- Specific sequences enabling intervention

**THAUMIEL:**
- Method paper (parameter-free scRNA-seq)
- Hormesis protocol (fasting, frequency, cold)
- Error-correction restoration approaches

### Tagging

In any document, tag inline: `[SAFE]` `[EUCLID]` `[KETER]` `[APOLLYON]` `[THAUMIEL]`

## Companion Protocols

| Protocol | Role | Description |
|----------|------|-------------|
| **HALO** | Memory | Living documents. Honest, Accessible, Living, Ours. |
| **WING** | Release gate | Working, Important, Novel, Good. All four YES to publish. |
| **EYE** | Perception | Observation and analysis layer. How we look at data. |
| **DAEMON** | Mind | Reads HALOs, checks WINGs, runs EYEs, decides what to do. |
  halo.created: auto
---
vault_clearance: EUCLID
  halo.created: auto
# HALO Protocol
> **H**onest. **A**ccessible. **L**iving. **O**urs.

## What This Is

Every entity in the lab — a cell culture, a dataset, a theory, a tool, a finding — gets a **Living Document** that follows a standardized structure but never stops growing. The document IS the entity's life story.

## The Living Document Format

Each Living Document has two layers:

### Layer 1: The Paper (public, growing)

| Section | What goes here | When it updates |
|---------|---------------|-----------------|
| **Abstract** | One paragraph. What this is, why it matters, what it showed. Pointers to lineage (parent experiments, related entities). | Every major finding |
| **Methods** | How it functions, behaves. Exact protocol, parameters, code paths. | When protocol changes |
| **Data** | Engines run, records, databases, raw observations. File locations, sizes, formats. | Every new data point |
| **Discussion** | Interpretation. Links to WORLDLINE sections where findings were analyzed. | After each analysis |
| **Conclusion** | What it means for the lab. What it means for the world. | When understanding deepens |
| **Next Steps** | Bounty board listings. New questions generated. | Continuously |

### Layer 2: The Containment (operational, security)

| Section | What goes here |
|---------|---------------|
| **Classification** | How sensitive is this? PUBLIC (can share), INTERNAL (lab only), RESTRICTED (key personnel), CLASSIFIED (encrypted) |
| **Containment** | What do we keep secret? What's the output format? What gets shared vs held? |
| **Addenda** | Append-only observations. Anomalies. Things that don't fit. "We noticed X but don't know what it means yet." |
| **Breach Protocol** | What happens if containment fails? If the data leaks? If someone scoops us? Contingency plan. |

## The Four Principles

### H — Honest
- State what we know AND what we don't
- Flag confidence levels (HIGH, MEDIUM, LOW, SPECULATIVE)
- Document failures and dead ends, not just successes
- If a dataset has caveats (P5-8 HUVECs, snRNA-seq missing mito), say so
- Never claim more than the data supports

### A — Accessible
- Any AI agent reading the project **BOOK** (and any registry subsection, e.g. §5) should understand the project in 30 seconds
- Any human collaborator should be able to run the pipeline in 3 commands
- File names describe their contents (no `data_final_v2_REAL.h5ad`)
- Living Documents link to each other through pointers, not assumptions
- The structure is self-documenting — the folder hierarchy IS the ontology

### L — Living
- Documents grow by APPENDING, not rewriting history
- Addenda are timestamped and signed (who observed this, when)
- The worldline is append-only — what happened stays in the record
- New findings update the Abstract and Conclusion but don't erase the old ones
- Dead ends get documented, not deleted — they're data too

### O — Ours
- The protocol belongs to the lab, not to any platform
- Works on any machine, any OS, any editor
- No vendor lock-in (GitHub, Notion, Obsidian are TOOLS, not dependencies)
- The encryption is semantic — the naming conventions ARE the code
- New collaborators earn access through contribution, not credentials

## File Structure Under HALO

```
Project/
  BOOK.md                # Bibliography + methods (+ fused registry subsection, e.g. §5 Combined Codex)
  QUICK_START.md         # 3-command onboarding
  README.md              # Public-facing overview
  WORLDLINE.md  # Append-only discovery log
  BOUNTY_BOARD.md        # Open questions and tasks
  HALO_PROTOCOL.md       # This file (living doc template: vault README § Living document template)

  methods/               # Code — the engines
    run_pipeline.py      # Main entry point
    *.py                 # Analysis modules

  data/
    experiments/         # Each experiment = one Living Document
      EXP01_.../
        LIVING_DOCUMENT.md
        DATA_LOCATION.txt
      EXP02_.../
        ...
    imaging/             # Microscopy — organized by condition + date
      YYYY-MM-DD_CellType_Condition/
        img_NNN.jpg
        vid_NNN.avi
    pathways/            # Gene set databases (.gmt)
    plots/               # Generated figures

  orthodox/              # Standard pipeline comparison (self-contained)
```

## How to Create a New Living Document

When you start a new experiment, dataset analysis, or theoretical exploration:

1. Create the folder: `data/experiments/EXPNN_descriptive_name/`
2. Start from the vault canonical template: [README § Living document template (HALO)](../README.md#living-document-template-halo), or copy a skeleton from [99_Archive/.../LIVING_DOCUMENT_TEMPLATE_symphony_variant.md](../99_Archive/root_reports/2026-03/unification_documentation_hardening/LIVING_DOCUMENT_TEMPLATE_symphony_variant.md) into `LIVING_DOCUMENT.md`
3. Fill in the Abstract (what are you doing and why)
4. Fill in Methods (exact protocol)
5. Set the Classification level
6. Add a pointer or row in **`BOOK.md`** (registry subsection where the project keeps one, e.g. §5)

Everything else grows organically as results come in.

## How to Read the Lab

1. **Start with `BOOK.md` / README** — find what exists (registry subsection if applicable)
2. **Read the Living Document** for any entity you care about
3. **Check WORLDLINE.md** for the analysis history
4. **Check BOUNTY_BOARD.md** for open questions
5. **Run the pipeline** if you want to verify anything

## The Worldline

The WORLDLINE.md file is the worldline — the append-only history of everything discovered. It never gets rewritten. New parts get added. Old parts stay. The worldline IS the lab's memory.

The Addenda in each Living Document are local worldlines — the history of that specific entity. Together, they form the complete record.

## Classification Tiers

| Tier | Symbol | Meaning | Sharing Rule | Effort to Protect |
|------|--------|---------|-------------|-------------------|
| **SAFE** | `[S]` | Show anyone. Public knowledge. | GitHub, Zenodo, papers | None |
| **EUCLID** | `[E]` | Uncertain response. Needs framing. | Publish with defensive context | Review before sharing |
| **KETER** | `[K]` | Danger to the lab if leaked. | Vault only. Core team. | Active containment |
| **APOLLYON** | `[A]` | Danger to world if leaked without defense. | Internal only. Defense published FIRST. | Maximum containment |
| **THAUMIEL** | `[T]` | Benefit to world. Should be shared. | Publish when WING criteria met. | Release when ready |

### Classification Flow

```
KETER (new finding)
  -> understand it -> EUCLID (theory with context)
  -> build defense -> THAUMIEL (cure/tool ready to share)
  -> attack surface -> APOLLYON (until defense published)
  -> defense published -> downgrade attack to EUCLID
  -> fully public -> SAFE
```

### Current Project Classifications

**SAFE:**
- Pipeline code (`methods/*.py`)
- Public dataset analyses (coupling hierarchy, cross-species)
- Project documentation (README, BOOK, QUICK_START)

**EUCLID:**
- Desync cascade theory
- Operator simplex theory
- Three-kingdom structure
- GBM operator victory state analysis
- Cross-species coupling hierarchy

**KETER:**
- All wet lab work, protocols, experimental designs
- iPSC coculture observations (`data/imaging/`)
- lncRNA checkpoint discovery (specific gene IDs)
- LIANA communication map (specific interactions)
- BAM file analyses
- Unpublished experimental results

**APOLLYON:**
- Chr19 UHRF1-DNMT1 TE insertion map (BloodyEchoes)
- Viral operator coupling analysis
- Synthetic TE insertion approach
- Specific sequences enabling intervention

**THAUMIEL:**
- Method paper (parameter-free scRNA-seq)
- Hormesis protocol (fasting, frequency, cold)
- Error-correction restoration approaches

### Tagging

In any document, tag inline: `[SAFE]` `[EUCLID]` `[KETER]` `[APOLLYON]` `[THAUMIEL]`

## Companion Protocols

| Protocol | Role | Description |
|----------|------|-------------|
| **HALO** | Memory | Living documents. Honest, Accessible, Living, Ours. |
| **WING** | Release gate | Working, Important, Novel, Good. All four YES to publish. |
| **EYE** | Perception | Observation and analysis layer. How we look at data. |
| **DAEMON** | Mind | Reads HALOs, checks WINGs, runs EYEs, decides what to do. |