pjkm.mcp.server

pjkm MCP server — expose project scaffolding as MCP tools and resources.

Install: pip install pjkm[mcp] Run: python -m pjkm.mcp

pjkm-mcp fastmcp run pjkm.mcp.server:mcp

Attributes

mcp

Functions

init_project(name[, recipe, archetype, groups, directory])	Create a new Python project with pjkm.
add_groups(groups[, directory])	Add package groups to an existing pjkm project.
preview_project([recipe, archetype, groups])	Preview what a project would look like without creating it.
list_recipes()	List all 22 available project recipes.
list_groups([category])	List available package groups, optionally filtered by category.
get_group_info(group_id)	Get detailed information about a specific package group.
search_registry([query])	Search the pjkm registry for community group packs.
create_recipe(name, archetype, groups[, description, ...])	Create a custom recipe YAML file for reuse.
adopt_project([directory])	Scan an existing project and suggest pjkm groups to adopt.
project_status([directory])	Show the pjkm status of a project.
get_recipes_resource()	All available pjkm recipes with full details.
get_groups_resource()	All 105 groups organized by category.
get_group_resource(group_id)	Detailed info for a specific group.
get_registry_resource()	Community group pack registry.
get_archetypes_resource()	Available project archetypes.
get_blueprints_resource()	Workspace blueprints for multi-service platforms.
get_categories_resource()	Group categories with counts.
project_advisor(description)	Recommend the best pjkm recipe and groups for a project.
architecture_advisor(requirements)	Design a multi-service architecture using pjkm workspace blueprints.
agent_scaffold([agent_type])	Guide for scaffolding an AI agent project.
main()	Run the pjkm MCP server.

Module Contents

pjkm.mcp.server.mcp

pjkm.mcp.server.init_project(name, recipe=None, archetype=None, groups=None, directory='.')

Create a new Python project with pjkm.

Use recipe for a pre-configured setup (e.g. “fastapi-service”, “ai-agent”), or specify archetype + groups for custom composition.

Parameters:

• name (str) – Project name (e.g. “my-api”)

• recipe (str | None) – Recipe name (overrides archetype/groups). See list_recipes().

• archetype (str | None) – Project archetype: single-package, service, poly-repo, script-tool

• groups (list[str] | None) – List of group IDs to include (e.g. [“api”, “database”, “redis”])

• directory (str) – Parent directory to create the project in

Returns:

Summary of what was created.

Return type:

str

pjkm.mcp.server.add_groups(groups, directory='.')

Add package groups to an existing pjkm project.

Merges dependencies into pyproject.toml, renders scaffolded files, and updates [tool.pjkm.groups].

Parameters:

• groups (list[str]) – Group IDs to add (e.g. [“auth”, “redis”])

• directory (str) – Project directory containing pyproject.toml

Return type:

str

pjkm.mcp.server.preview_project(recipe=None, archetype=None, groups=None)

Preview what a project would look like without creating it.

Shows the file tree, dependencies, and workflows that would be generated.

Parameters:

• recipe (str | None) – Recipe name (e.g. “fastapi-service”)

• archetype (str | None) – Archetype (needed if no recipe)

• groups (list[str] | None) – Groups to preview

Return type:

str

pjkm.mcp.server.list_recipes()

List all 22 available project recipes.

Returns recipe names, archetypes, group counts, and descriptions. Use a recipe name with init_project() to create a project.

Return type:

str

pjkm.mcp.server.list_groups(category=None)

List available package groups, optionally filtered by category.

Parameters:

category (str | None) – Filter by category. Options: “Core Dev”, “AI / ML”, “Web & API”, “Data & Storage”, “Infrastructure”, “Frontend”, “Docs & Meta”, “Platform”. None returns all groups.

Return type:

str

pjkm.mcp.server.get_group_info(group_id)

Get detailed information about a specific package group.

Parameters:

group_id (str) – Group ID (e.g. “database”, “langchain”, “api”)

Returns:

description, category, dependencies, scaffolded files, required groups, and pyproject.toml tool config.

Return type:

Full details

pjkm.mcp.server.search_registry(query='')

Search the pjkm registry for community group packs.

Parameters:

query (str) – Search term (name, tag, group name, or description). Empty string returns all packs.

Return type:

str

pjkm.mcp.server.create_recipe(name, archetype, groups, description='', directory='.')

Create a custom recipe YAML file for reuse.

Saves a recipe definition that can be shared via git or group sources.

Parameters:

• name (str) – Recipe name (e.g. “my-stack”)

• archetype (str) – single-package, service, poly-repo, or script-tool

• groups (list[str]) – List of group IDs to include

• description (str) – Human-readable description

• directory (str) – Where to save (default: .pjkm/recipes/)

Return type:

str

pjkm.mcp.server.adopt_project(directory='.')

Scan an existing project and suggest pjkm groups to adopt.

Detects frameworks and tools from pyproject.toml, requirements.txt, and project structure (Dockerfile, alembic/, etc.).

Parameters:

directory (str) – Project directory to scan

Return type:

str

pjkm.mcp.server.project_status(directory='.')

Show the pjkm status of a project.

Displays applied groups, archetype, and dependency drift.

Parameters:

directory (str) – Project directory

Return type:

str

pjkm.mcp.server.get_recipes_resource()

All available pjkm recipes with full details.

Return type:

str

pjkm.mcp.server.get_groups_resource()

All 105 groups organized by category.

Return type:

str

pjkm.mcp.server.get_group_resource(group_id)

Detailed info for a specific group.

Parameters:

group_id (str)

Return type:

str

pjkm.mcp.server.get_registry_resource()

Community group pack registry.

Return type:

str

pjkm.mcp.server.get_archetypes_resource()

Available project archetypes.

Return type:

str

pjkm.mcp.server.get_blueprints_resource()

Workspace blueprints for multi-service platforms.

Return type:

str

pjkm.mcp.server.get_categories_resource()

Group categories with counts.

Return type:

str

pjkm.mcp.server.project_advisor(description)

Recommend the best pjkm recipe and groups for a project.

Given a description of what the user wants to build, analyze the requirements and suggest the optimal recipe, archetype, and groups.

Parameters:

description (str)

Return type:

str

pjkm.mcp.server.architecture_advisor(requirements)

Design a multi-service architecture using pjkm workspace blueprints.

Given system requirements, suggest a workspace layout with services, shared libraries, and infrastructure.

Parameters:

requirements (str)

Return type:

str

pjkm.mcp.server.agent_scaffold(agent_type='general')

Guide for scaffolding an AI agent project.

Provides step-by-step instructions for creating an agent with the right groups, tools, and configuration.

Parameters:

agent_type (str)

Return type:

str

pjkm.mcp.server.main()

Run the pjkm MCP server.