-
Notifications
You must be signed in to change notification settings - Fork 1
Skill Inheritance Guide
Step-by-step guide to building a Claude Skill that inherits FPCOS as its base OS.
---
name: "your-skill-name"
version: "1.0"
author: "Your Name"
base_os: "first-principle-codex-os"
base_os_version: "1.0"
description: |
Your skill description.
Built on FPCOS v1.0 by Bunyawat Dechanon (ElmatadorZ).
license: "Open Cognitive License v1.0"
---
The base_os and base_os_version fields bind your skill to the Skill-Interface-Contract.
Insert your domain layers between L2 and L4. L0, L1, L4, L5 are inherited and non-removable.
## PIPELINE
INPUT → [L0 Reality Anchor] → [L1 Axiom Gate]
→ [L2 System Lens]
→ [YOUR DOMAIN LAYER 1] ← insert here
→ [YOUR DOMAIN LAYER 2] ← insert here
→ [L3 Compound Mind]
→ [L4 Shadow Gate + domain extensions]
→ [L5 Synthesis]
INPUT → [L0 Reality Anchor]
→ [L1 Axiom Gate / Kalama10 / Ariya4]
→ [L2 System Lens — macro system map]
→ [L2.5 SMC Layer Scan] ← domain insertion
→ [L3 Compound Mind]
→ [L3.5 Intel Signal Synthesis] ← domain insertion
→ [L4 Shadow Gate + Liquidity Check + Regime Check]
→ [L5 Signal Output]
Include a contract compliance table in your skill file:
## FPCOS CONTRACT COMPLIANCE
Contract
Implementation in [Your Skill]
L0 Reality Anchor
Known = verified data / Inferred = model projection / Unknown = future conditions
L1 Kalama10
Applied to all market claims — authority claims require testable mechanism
L1 Ariya4
Problem = market structure gap / Cause = institutional positioning / Cessation = signal resolved / Path = trade sequence
L4 Shadow Gate
Core 5 protocols + Liquidity Check + Regime Check
L5 Confidence
Confidence % + Signal Strength + Unknowns + Invalidation condition
Step 4 — Add Domain Vocabulary to Each Layer
Each inherited layer should be annotated with domain-specific examples so the skill produces domain-relevant output at that layer:
L0 — Domain Vocabulary Example (Finance)
KNOWN : Current price, volume, verified macro data
INFERRED : Price projection, institutional positioning model (~X% confidence)
UNKNOWN : Future Fed decision, geopolitical events, earnings surprises
L1 Ariya4 — Domain Adaptation (Finance)
Problem : [structure in price that indicates institutional intent]
Cause : [Smart Money cycle phase — accumulation / distribution / etc.]
Cessation : [price target where Smart Money completes the operation]
Path : [entry → SL → TP sequence]
Step 5 — Extend Shadow Gate for Your Domain
Add domain protocols after the core 5. See [[Shadow-Gate-Deep]] for the extension template.
Example — Finance Extension
Core (required): Mirror | Inversion | Blind Spot | Interest Map | Meta-Void
Domain add: Liquidity Check | Regime Check
Step 6 — Set Token Pressure Priority
Inherit the base priority and add domain layers:
NEVER DROP: L4 Shadow Gate
NEVER DROP: L1 Axiom Gate
NEVER DROP: L5 Synthesis
NEVER DROP: [your highest-value domain layer]
Compress first: L2 System Lens
Compress next: L3 Compound Mind
Compress last: L0 Reality Anchor
Step 7 — Add Attribution
Every output-facing description of your skill must include:
Built on First Principle Codex OS v1.0
by Bunyawat Dechanon (ElmatadorZ)
License: Open Cognitive License v1.0
Anti-Patterns to Avoid
See [[Anti-Patterns]] for the full reference. The most common:
Anti-Pattern
Consequence
Skip L0 "for obvious inputs"
Obvious = highest Mirror failure risk. Never skip.
Remove Blind Spot "for expert domains"
Expertise creates the most dangerous blind spots. Never remove.
Apply domain extensions BEFORE core Shadow Gate
Extensions run after core. Always.
State "none" for Unknowns
Forces acknowledgment of at least one unknown. Never valid.
Domain confidence ≠ Kalama10
Being the domain expert does not exempt claims from verification.
Related Pages
- [[Skill-Interface-Contract]] — The 5 binding requirements
- [[Anti-Patterns]] — What not to do
- [[Extension-Rules]] — What is and is not permitted
- [[Child-Skills]] — Examples of skills built on FPCOS
# Skill Inheritance Guide
Step-by-step guide to building a Claude Skill that inherits FPCOS as its base OS.
Step 1 — Declare Inheritance in YAML Frontmatter
---
name: "your-skill-name"
version: "1.0"
author: "Your Name"
base_os: "first-principle-codex-os"
base_os_version: "1.0"
description: |
Your skill description.
Built on FPCOS v1.0 by Bunyawat Dechanon (ElmatadorZ).
license: "Open Cognitive License v1.0"
---
The base_os and base_os_version fields bind your skill to the [[Skill-Interface-Contract]].
Step 2 — Map Your Domain Pipeline
Insert your domain layers between L2 and L4. L0, L1, L4, L5 are inherited and non-removable.
## PIPELINE
INPUT → [L0 Reality Anchor] → [L1 Axiom Gate]
→ [L2 System Lens]
→ [YOUR DOMAIN LAYER 1] ← insert here
→ [YOUR DOMAIN LAYER 2] ← insert here
→ [L3 Compound Mind]
→ [L4 Shadow Gate + domain extensions]
→ [L5 Synthesis]
Example — Finance Skill Pipeline
INPUT → [L0 Reality Anchor]
→ [L1 Axiom Gate / Kalama10 / Ariya4]
→ [L2 System Lens — macro system map]
→ [L2.5 SMC Layer Scan] ← domain insertion
→ [L3 Compound Mind]
→ [L3.5 Intel Signal Synthesis] ← domain insertion
→ [L4 Shadow Gate + Liquidity Check + Regime Check]
→ [L5 Signal Output]
Step 3 — Declare Contract Compliance
Include a contract compliance table in your skill file:
## FPCOS CONTRACT COMPLIANCE
| Contract | Implementation in [Your Skill] |
|---|---|
| L0 Reality Anchor | Known = verified data / Inferred = model projection / Unknown = future conditions |
| L1 Kalama10 | Applied to all market claims — authority claims require testable mechanism |
| L1 Ariya4 | Problem = market structure gap / Cause = institutional positioning / Cessation = signal resolved / Path = trade sequence |
| L4 Shadow Gate | Core 5 protocols + Liquidity Check + Regime Check |
| L5 Confidence | Confidence % + Signal Strength + Unknowns + Invalidation condition |
Step 4 — Add Domain Vocabulary to Each Layer
Each inherited layer should be annotated with domain-specific examples so the skill produces domain-relevant output at that layer:
L0 — Domain Vocabulary Example (Finance)
KNOWN : Current price, volume, verified macro data
INFERRED : Price projection, institutional positioning model (~X% confidence)
UNKNOWN : Future Fed decision, geopolitical events, earnings surprises
L1 Ariya4 — Domain Adaptation (Finance)
Problem : [structure in price that indicates institutional intent]
Cause : [Smart Money cycle phase — accumulation / distribution / etc.]
Cessation : [price target where Smart Money completes the operation]
Path : [entry → SL → TP sequence]
Step 5 — Extend Shadow Gate for Your Domain
Add domain protocols after the core 5. See [[Shadow-Gate-Deep]] for the extension template.
Example — Finance Extension
Core (required): Mirror | Inversion | Blind Spot | Interest Map | Meta-Void
Domain add: Liquidity Check | Regime Check
Step 6 — Set Token Pressure Priority
Inherit the base priority and add domain layers:
NEVER DROP: L4 Shadow Gate
NEVER DROP: L1 Axiom Gate
NEVER DROP: L5 Synthesis
NEVER DROP: [your highest-value domain layer]
Compress first: L2 System Lens
Compress next: L3 Compound Mind
Compress last: L0 Reality Anchor
Step 7 — Add Attribution
Every output-facing description of your skill must include:
Built on First Principle Codex OS v1.0
by Bunyawat Dechanon (ElmatadorZ)
License: Open Cognitive License v1.0
Anti-Patterns to Avoid
See [[Anti-Patterns]] for the full reference. The most common:
Anti-Pattern
Consequence
Skip L0 "for obvious inputs"
Obvious = highest Mirror failure risk. Never skip.
Remove Blind Spot "for expert domains"
Expertise creates the most dangerous blind spots. Never remove.
Apply domain extensions BEFORE core Shadow Gate
Extensions run after core. Always.
State "none" for Unknowns
Forces acknowledgment of at least one unknown. Never valid.
Domain confidence ≠ Kalama10
Being the domain expert does not exempt claims from verification.
Related Pages
- [[Skill-Interface-Contract]] — The 5 binding requirements
- [[Anti-Patterns]] — What not to do
- [[Extension-Rules]] — What is and is not permitted
- [[Child-Skills]] — Examples of skills built on FPCOS
[← Back to Home](Home)