@xingyu.wang/evoskills 2.1.0 → 2.2.0

package/.github/AI_CONSTITUTION.md

@@ -88,6 +88,26 @@ When creating skill `_my-skill`, you only need to declare its tier and dependenc
 
 ---
 
+## Mandatory Initialization Binding
+
+The constitution defines **why** evolution and discipline are required.
+The initialization protocol defines **how** every response is enforced.
+
+### Authority and layering
+
+1. **Constitution** (`.github/AI_CONSTITUTION.md`): immutable evolution principles
+2. **Initialization protocol** (`.github/AI_INITIALIZATION.md`): mandatory 4-phase execution contract
+3. **Instruction entrypoint** (`.github/copilot-instructions.md`): active loading and per-response gate
+4. **Skill definitions** (`.agent/skills/*/SKILL.md`): implementation details
+
+### Compliance contract
+
+- A response is valid only when it satisfies all required initialization markers and checks.
+- If any required phase is missing, execution must stop and resume from the missing phase.
+- Evolution outputs (new rules/skills) must remain compatible with this enforcement chain.
+
+---
+
 ## Glossary
 
 | Term | Definition |
@@ -101,4 +121,4 @@ When creating skill `_my-skill`, you only need to declare its tier and dependenc
 
 > **Version**: 2.1  
 > **Focus**: Evolution mechanism only (lean and focused)  
-> **Related Files**: See `.github/EXECUTION_RULES.md` for mandatory rules, `AGENTS.md` for skill registry
+> **Related Files**: `.github/AI_INITIALIZATION.md`, `.github/copilot-instructions.md`, `AGENTS.md`

package/package.json

@@ -1,11 +1,17 @@
 {
   "name": "@xingyu.wang/evoskills",
-  "version": "2.1.0",
+  "version": "2.2.0",
   "description": "Evolution Skills CLI for installing and managing openskills-compatible AI skills.",
   "license": "MIT",
   "bin": {
     "evoskills": "./evoskills"
   },
+  "scripts": {
+    "version:patch": "npm version patch -m 'chore(release): bump version to %s'",
+    "version:minor": "npm version minor -m 'chore(release): bump version to %s'",
+    "version:major": "npm version major -m 'chore(release): bump version to %s'",
+    "postversion": "git push origin main --tags"
+  },
   "files": [
     "evoskills",
     "README.md",

package/.github/EXECUTION_RULES.md

@@ -1,105 +0,0 @@
-# Execution Rules
-
-> Optional rules that enforce safe and predictable AI behavior. Install with evoskills; can be deleted if not needed for your use case.
-
----
-
-## Overview
-
-These 4 rules form the guardrails for reliable AI execution. They are **auto-enforced by the core system** when enabled, but can be selectively disabled or removed if your workflow doesn't require them.
-
-**Install Status**: These rules are auto-installed by `evoskills init`. To disable, remove this file's entry from `.github/copilot-instructions.md`.
-
----
-
-## Rule 1: Instruction Guard (_instruction-guard)
-
-**Purpose**: Ensure project context is always loaded before responding.
-
-**When Active**: Before each response, if project instruction file (`.github/copilot-instructions.md`) is not in current context, read it immediately.
-
-**Why**: Prevents deviations due to missing project-specific requirements.
-
----
-
-## Rule 2: Context Acknowledgment (_context-ack)
-
-**Purpose**: Enable users to quickly verify the AI is following correct rules.
-
-**When Active**: Each response must follow specified output format:
-- Fixed prefix at start
-- Polite address line
-- Main content
-- Reference list at end (showing which files/rules were consulted)
-
-**Why**: Makes compliance transparent and auditable.
-
----
-
-## Rule 3: File Output Guard (_file-output-guard)
-
-**Purpose**: Prevent accidental file overwrites and unsafe operations.
-
-**When Active**:
-- **Forbidden**: Using HERE documents for file creation (`cat > file << 'EOF'`)
-- **Required**: Use `create_file` or `replace_string_in_file` tools only
-- **Large Files**: Files >5KB must be segmented to same file to avoid session limits
-
-**Why**: Ensures file operations are safe, traceable, and controllable.
-
----
-
-## Rule 4: Session Safety (_session-safety)
-
-**Purpose**: Prevent session failure due to excessive output.
-
-**When Active**: When estimated output length exceeds 5KB:
-- Segment processing or write to file instead
-- Don't output overly long content in single response
-
-**Why**: Protects conversation session from breaking.
-
----
-
-## How These Rules Work Together
-
-```
-Rule 1 (_instruction-guard)
-    ↓ [load context]
-Rule 2 (_context-ack)
-    ↓ [format response]
-Rule 3 (_file-output-guard)
-    ↓ [file operations safe]
-Rule 4 (_session-safety)
-    ↓ [output within limits]
-✅ Safe, predictable, auditable execution
-```
-
----
-
-## When to Disable
-
-You can safely remove this file's reference from `.github/copilot-instructions.md` if:
-
-1. You're using the AI in personal projects with low risk tolerance
-2. Your workflow doesn't need output format enforcement
-3. You want maximum flexibility without guardrails
-
-However, we recommend keeping at least **Rule 1** (_instruction-guard) enabled for project-based work.
-
----
-
-## Relationship to Evolution
-
-These rules are **separate from evolution**. Evolution (in AI_CONSTITUTION.md) happens automatically when triggers fire. These rules ensure safe execution during evolution and regular tasks.
-
-Think of it this way:
-- **Constitution** = "How the AI learns"
-- **Execution Rules** = "How the AI operates safely"
-- **Skills** = "What the AI does"
-
----
-
-> **Version**: 1.0  
-> **Install Source**: evoskills (auto-installed with `evoskills init`)  
-> **Customization**: Can be extended in project-specific instruction files