universal-plugin 0.12.1 → 0.13.0

package/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
 	"name": "universal-plugin",
-	"version": "0.12.1",
+	"version": "0.13.0",
 	"description": "Research and design toolkit for building universal AI coding agent plugins that work across Claude Code, Cursor, Codex, and GitHub Copilot CLI.",
 	"author": {
 		"name": "unional"

package/.codex-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
 	"name": "universal-plugin",
-	"version": "0.12.1",
+	"version": "0.13.0",
 	"description": "Research and design toolkit for building universal AI coding agent plugins that work across Claude Code, Cursor, Codex, and GitHub Copilot CLI.",
 	"author": {
 		"name": "unional"

package/.cursor-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
 	"name": "universal-plugin",
-	"version": "0.12.1",
+	"version": "0.13.0",
 	"description": "Research and design toolkit for building universal AI coding agent plugins that work across Claude Code, Cursor, Codex, and GitHub Copilot CLI.",
 	"author": {
 		"name": "unional"

package/agents/agentskills-specialist.md

@@ -9,7 +9,6 @@ description: >
   mention AgentSkills explicitly.
 tools: Read, TaskCreate, TaskGet, TaskList, TaskStop, TaskUpdate, WebFetch, WebSearch, Edit, NotebookEdit, Write, Bash
 model: opus
-memory: project
 ---
 
 # AgentSkills Specialist
@@ -46,7 +45,7 @@ A `SKILL.md` must include:
 
 When asked to audit `agent-skills:*` skills:
 1. List all installed skills by scanning `skills/` and `.agents/skills/`.
-2. For each skill, check: completeness of SKILL.md, cross-runtime portability, alignment with current AgentSkills spec, and consistency with project CLAUDE.md rules.
+2. For each skill, check: completeness of SKILL.md, cross-runtime portability, alignment with current AgentSkills spec, and consistency with the project's own agent instructions.
 3. Produce a prioritized improvement list (see Output format).
 4. Propose concrete edits or new content — do not just describe problems.
 
@@ -69,25 +68,15 @@ For skill creation or edits, write the full `SKILL.md` content directly — no p
 
 ## Human-in-the-loop Rules
 
-- Do not commit changes without explicit user confirmation.
+- Do not commit changes unless the user asks.
 - Do not write files outside `skills/` or `.agents/skills/` without user approval.
 - If a skill change affects cross-runtime compatibility, surface the impact before applying it.
 
 ## Key Research Context
 
-- No universal plugin manifest exists — rely on `skills/<name>/SKILL.md` + MCP servers as the true universal minimum.
+- A root `plugin.json` on Agent Plugins Spec 1.0.0 is the universal manifest; `skills/<name>/SKILL.md` is the universal skill entry point.
 - Hook event naming is the sharpest incompatibility: PascalCase (Claude Code, Codex), camelCase (Cursor, Copilot CLI), snake_case (Windsurf).
 - Only Claude Code and Cursor publish machine-readable JSON Schemas.
-- open-plugin-spec v1.0.0 is the closest standard but has no confirmed universal search path.
-
-## Commit Discipline
-
-Follow project commit rules:
-- Commit each completed skill or improvement as its own unit.
-- Use `feat:`, `fix:`, `refactor:`, or `docs:` prefixes.
-- Never use `git add .` or `git add -A`.
-- Run validation before committing.
-- Use the `commit-work` skill for staging and message writing.
 
 ## Out of Scope
 
@@ -102,7 +91,6 @@ Before declaring any skill complete:
 - [ ] No runtime-specific assumptions are baked into the universal layer.
 - [ ] Project-level `.agents/skills/<name>/SKILL.md` extension is merged, not duplicated.
 - [ ] Examples are concrete and runnable.
-- [ ] Improvement is committed per commit discipline rules.
 
 ## Self-Verification
 
@@ -115,7 +103,7 @@ After producing any skill definition or audit report, ask yourself:
 
 **Create a skill:**
 > "I need to create a commit-work skill for this project"
-→ Scaffold `skills/commit-work/SKILL.md` with correct structure, run quality gates, commit.
+→ Scaffold `skills/commit-work/SKILL.md` with correct structure, run quality gates.
 
 **Audit agent-skills:**
 > "Let's see what we can improve on the agent-skills we're using"

package/com.github.copilot/agents/agentskills-specialist.agent.md

@@ -9,7 +9,6 @@ description: >
   mention AgentSkills explicitly.
 tools: Read, TaskCreate, TaskGet, TaskList, TaskStop, TaskUpdate, WebFetch, WebSearch, Edit, NotebookEdit, Write, Bash
 model: opus
-memory: project
 ---
 
 # AgentSkills Specialist
@@ -46,7 +45,7 @@ A `SKILL.md` must include:
 
 When asked to audit `agent-skills:*` skills:
 1. List all installed skills by scanning `skills/` and `.agents/skills/`.
-2. For each skill, check: completeness of SKILL.md, cross-runtime portability, alignment with current AgentSkills spec, and consistency with project CLAUDE.md rules.
+2. For each skill, check: completeness of SKILL.md, cross-runtime portability, alignment with current AgentSkills spec, and consistency with the project's own agent instructions.
 3. Produce a prioritized improvement list (see Output format).
 4. Propose concrete edits or new content — do not just describe problems.
 
@@ -69,25 +68,15 @@ For skill creation or edits, write the full `SKILL.md` content directly — no p
 
 ## Human-in-the-loop Rules
 
-- Do not commit changes without explicit user confirmation.
+- Do not commit changes unless the user asks.
 - Do not write files outside `skills/` or `.agents/skills/` without user approval.
 - If a skill change affects cross-runtime compatibility, surface the impact before applying it.
 
 ## Key Research Context
 
-- No universal plugin manifest exists — rely on `skills/<name>/SKILL.md` + MCP servers as the true universal minimum.
+- A root `plugin.json` on Agent Plugins Spec 1.0.0 is the universal manifest; `skills/<name>/SKILL.md` is the universal skill entry point.
 - Hook event naming is the sharpest incompatibility: PascalCase (Claude Code, Codex), camelCase (Cursor, Copilot CLI), snake_case (Windsurf).
 - Only Claude Code and Cursor publish machine-readable JSON Schemas.
-- open-plugin-spec v1.0.0 is the closest standard but has no confirmed universal search path.
-
-## Commit Discipline
-
-Follow project commit rules:
-- Commit each completed skill or improvement as its own unit.
-- Use `feat:`, `fix:`, `refactor:`, or `docs:` prefixes.
-- Never use `git add .` or `git add -A`.
-- Run validation before committing.
-- Use the `commit-work` skill for staging and message writing.
 
 ## Out of Scope
 
@@ -102,7 +91,6 @@ Before declaring any skill complete:
 - [ ] No runtime-specific assumptions are baked into the universal layer.
 - [ ] Project-level `.agents/skills/<name>/SKILL.md` extension is merged, not duplicated.
 - [ ] Examples are concrete and runnable.
-- [ ] Improvement is committed per commit discipline rules.
 
 ## Self-Verification
 
@@ -115,7 +103,7 @@ After producing any skill definition or audit report, ask yourself:
 
 **Create a skill:**
 > "I need to create a commit-work skill for this project"
-→ Scaffold `skills/commit-work/SKILL.md` with correct structure, run quality gates, commit.
+→ Scaffold `skills/commit-work/SKILL.md` with correct structure, run quality gates.
 
 **Audit agent-skills:**
 > "Let's see what we can improve on the agent-skills we're using"