Agent Plugin Packager Skill
SOP Type: Automated Packaging & Verification
Standard: Agent Plugins 1.0.0 Specification (agent-plugins.org/specification)
This skill defines the procedural Standard Operating Procedure (SOP) for packaging DSOM skills, spatial memory, and MCP servers into vendor-neutral Agent Plugins 1.0.0 packages.
🧭 Operational Flow
flowchart TD
Identify["1. Identify Skill & MCP Components"] --> Validate["2. Validate OKF v0.2 & Path Containment"]
Validate --> Scaffold["3. Generate plugin.json & mcp.json"]
Scaffold --> Bundle["4. Copy/Symlink to skills/ Directory"]
Bundle --> Verify["5. Run Conformance Test Suite"]
🛠️ Step-by-Step Instructions
Step 1: Scaffold the Plugin Package Directory
# PowerShell
$PluginName = "dsom-sysops-plugin"
mkdir $PluginName
mkdir $PluginName/skills
# Bash
export PLUGIN_NAME="dsom-sysops-plugin"
mkdir -p "$PLUGIN_NAME/skills"
Step 2: Create Canonical plugin.json
Create $PLUGIN_NAME/plugin.json adhering to schema 1.0.0:
{
"$schema": "https://agent-plugins.org/schemas/1.0.0/plugin.schema.json",
"name": "dsom-sysops-plugin",
"version": "1.0.0",
"description": "System operations and diagnostic skills for AI Agents.",
"keywords": ["sysops", "diagnostics", "dsom"]
}
Step 3: Populate skills/ Directory
Copy target skills into $PLUGIN_NAME/skills/<skill-name>/:
Ensure each skill folder contains:
- SKILL.md (valid OKF v0.2 frontmatter)
- scripts/ (executable scripts)
- references/ (supporting docs)
Step 4: Configure mcp.json (If Tools are Included)
{
"$schema": "https://agent-plugins.org/schemas/1.0.0/mcp.schema.json",
"mcpServers": {
"server": {
"type": "stdio",
"command": "uv",
"args": ["run", "python", "${PLUGIN_ROOT}/tools/mcp/server.py"],
"cwd": "${PLUGIN_ROOT}"
}
}
}
Step 5: Validate Conformance
Run the DSOM agent plugins test suite:
uv run pytest tests/test_agent_plugins_spec.py
Deep State of Mind (DSOM) For My AI Protocol | Harisfazillah Jamel (LinuxMalaysia) | 2026-08-22
Standard: UK English | DBP-standard Bahasa Melayu Malaysia (Piawai) | GNU General Public License v3.0