agentskills-mcp-server 0.2.0__tar.gz

agentskills_mcp_server-0.2.0/PKG-INFO

@@ -0,0 +1,179 @@
+Metadata-Version: 2.4
+Name: agentskills-mcp-server
+Version: 0.2.0
+Summary: MCP server integration for the Agent Skills format — expose skills as MCP tools and resources (https://agentskills.io)
+License: MIT
+Author: Pratik Panda
+Requires-Python: >=3.12,<4.0
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
+Classifier: Topic :: Software Development :: Libraries
+Provides-Extra: fs
+Provides-Extra: http
+Requires-Dist: agentskills-core (>=0.1.0)
+Requires-Dist: mcp (>=1.0)
+Requires-Dist: pydantic (>=2.0)
+Project-URL: Homepage, https://agentskills.io
+Project-URL: Repository, https://github.com/pratikxpanda/agentskills-sdk
+Description-Content-Type: text/markdown
+
+# agentskills-mcp-server
+
+[![PyPI](https://img.shields.io/pypi/v/agentskills-mcp-server)](https://pypi.org/project/agentskills-mcp-server/)
+[![Python 3.12+](https://img.shields.io/pypi/pyversions/agentskills-mcp-server)](https://pypi.org/project/agentskills-mcp-server/)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://github.com/pratikxpanda/agentskills-sdk/blob/main/LICENSE)
+
+> MCP server integration for the [Agent Skills SDK](../../README.md) — expose a skill registry as an MCP server.
+
+Creates a [Model Context Protocol](https://modelcontextprotocol.io/) server from a `SkillRegistry`, exposing skills as MCP tools and resources. Works with any MCP-compatible client (Claude Desktop, VS Code, custom clients, etc.).
+
+## Installation
+
+```bash
+pip install agentskills-mcp-server
+```
+
+With provider extras:
+
+```bash
+pip install agentskills-mcp-server[fs]    # filesystem provider
+pip install agentskills-mcp-server[http]  # HTTP provider
+```
+
+Requires Python 3.12+. Installs `agentskills-core`, `mcp`, and `pydantic` as dependencies.
+
+## Quick Start (CLI)
+
+Create a `server.json` config file:
+
+```json
+{
+    "name": "My Skills Server",
+    "skills": [
+        {
+            "id": "incident-response",
+            "provider": "fs",
+            "options": {"root": "./skills"}
+        }
+    ]
+}
+```
+
+Start the server:
+
+```bash
+python -m agentskills_mcp_server --config server.json
+```
+
+With Streamable HTTP transport:
+
+```bash
+python -m agentskills_mcp_server --config server.json --transport streamable-http
+```
+
+The server listens on `http://127.0.0.1:8000/mcp`.
+
+### MCP Client Integration
+
+Any MCP-compatible client (Claude Desktop, VS Code, etc.) can connect to the server.
+
+Stdio (local):
+
+```json
+{
+    "command": "python",
+    "args": ["-m", "agentskills_mcp_server", "--config", "server.json"]
+}
+```
+
+Streamable HTTP (remote):
+
+```json
+{
+    "url": "http://127.0.0.1:8000/mcp"
+}
+```
+
+## Config Reference
+
+The `server.json` file supports the following structure:
+
+| Field | Type | Required | Description |
+| --- | --- | --- | --- |
+| `name` | `str` | Yes | Display name shown to MCP clients |
+| `instructions` | `str` | No | Server-level instructions sent during handshake |
+| `skills` | `list` | Yes | One or more skill definitions (see below) |
+
+Each skill entry:
+
+| Field | Type | Required | Description |
+| --- | --- | --- | --- |
+| `id` | `str` | Yes | Skill identifier |
+| `provider` | `str` | Yes | Provider type: `"fs"` or `"http"` |
+| `options` | `dict` | No | Provider-specific options |
+
+**Provider options:**
+
+- **`fs`**: `root` (path to skills directory, default `"."`)
+- **`http`**: `base_url` (required), `headers` (optional), `params` (optional query string parameters)
+
+## Programmatic Usage
+
+For custom providers or advanced setups, use the Python API directly:
+
+```python
+from agentskills_core import SkillRegistry
+from agentskills_mcp_server import create_mcp_server
+
+registry = SkillRegistry()
+await registry.register("incident-response", my_custom_provider)
+
+server = create_mcp_server(registry, name="My Skills Server")
+server.run()  # stdio by default
+```
+
+## Tools
+
+The server exposes tools that let the LLM agent access skill content:
+
+| Tool | Parameters | Description |
+| --- | --- | --- |
+| `get_skill_metadata` | `skill_id` | Read frontmatter (name, description, etc.) |
+| `get_skill_body` | `skill_id` | Load full skill instructions |
+| `get_skill_reference` | `skill_id`, `name` | Read a reference document |
+| `get_skill_script` | `skill_id`, `name` | Read a script |
+| `get_skill_asset` | `skill_id`, `name` | Read an asset |
+
+## Resources
+
+The server provides resources for system-prompt context:
+
+| URI | Description |
+| --- | --- |
+| `skills://catalog/xml` | XML catalog of all registered skills |
+| `skills://catalog/markdown` | Markdown catalog of all registered skills |
+| `skills://tools-usage-instructions` | Workflow instructions for using the tools |
+
+The MCP client reads these resources and injects them into the system prompt, giving the agent both *what* skills exist and *how* to interact with them.
+
+## API
+
+### `create_mcp_server(registry, *, name, instructions=None) -> FastMCP`
+
+| Parameter | Type | Description |
+| --- | --- | --- |
+| `registry` | `SkillRegistry` | The registry whose skills are exposed |
+| `name` | `str` | Display name for the MCP server (required) |
+| `instructions` | `str \| None` | Optional server-level instructions sent to clients |
+
+Returns a configured `FastMCP` instance ready for `server.run()`.
+
+## License
+
+MIT
+

agentskills_mcp_server-0.2.0/README.md

@@ -0,0 +1,154 @@
+# agentskills-mcp-server
+
+[![PyPI](https://img.shields.io/pypi/v/agentskills-mcp-server)](https://pypi.org/project/agentskills-mcp-server/)
+[![Python 3.12+](https://img.shields.io/pypi/pyversions/agentskills-mcp-server)](https://pypi.org/project/agentskills-mcp-server/)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://github.com/pratikxpanda/agentskills-sdk/blob/main/LICENSE)
+
+> MCP server integration for the [Agent Skills SDK](../../README.md) — expose a skill registry as an MCP server.
+
+Creates a [Model Context Protocol](https://modelcontextprotocol.io/) server from a `SkillRegistry`, exposing skills as MCP tools and resources. Works with any MCP-compatible client (Claude Desktop, VS Code, custom clients, etc.).
+
+## Installation
+
+```bash
+pip install agentskills-mcp-server
+```
+
+With provider extras:
+
+```bash
+pip install agentskills-mcp-server[fs]    # filesystem provider
+pip install agentskills-mcp-server[http]  # HTTP provider
+```
+
+Requires Python 3.12+. Installs `agentskills-core`, `mcp`, and `pydantic` as dependencies.
+
+## Quick Start (CLI)
+
+Create a `server.json` config file:
+
+```json
+{
+    "name": "My Skills Server",
+    "skills": [
+        {
+            "id": "incident-response",
+            "provider": "fs",
+            "options": {"root": "./skills"}
+        }
+    ]
+}
+```
+
+Start the server:
+
+```bash
+python -m agentskills_mcp_server --config server.json
+```
+
+With Streamable HTTP transport:
+
+```bash
+python -m agentskills_mcp_server --config server.json --transport streamable-http
+```
+
+The server listens on `http://127.0.0.1:8000/mcp`.
+
+### MCP Client Integration
+
+Any MCP-compatible client (Claude Desktop, VS Code, etc.) can connect to the server.
+
+Stdio (local):
+
+```json
+{
+    "command": "python",
+    "args": ["-m", "agentskills_mcp_server", "--config", "server.json"]
+}
+```
+
+Streamable HTTP (remote):
+
+```json
+{
+    "url": "http://127.0.0.1:8000/mcp"
+}
+```
+
+## Config Reference
+
+The `server.json` file supports the following structure:
+
+| Field | Type | Required | Description |
+| --- | --- | --- | --- |
+| `name` | `str` | Yes | Display name shown to MCP clients |
+| `instructions` | `str` | No | Server-level instructions sent during handshake |
+| `skills` | `list` | Yes | One or more skill definitions (see below) |
+
+Each skill entry:
+
+| Field | Type | Required | Description |
+| --- | --- | --- | --- |
+| `id` | `str` | Yes | Skill identifier |
+| `provider` | `str` | Yes | Provider type: `"fs"` or `"http"` |
+| `options` | `dict` | No | Provider-specific options |
+
+**Provider options:**
+
+- **`fs`**: `root` (path to skills directory, default `"."`)
+- **`http`**: `base_url` (required), `headers` (optional), `params` (optional query string parameters)
+
+## Programmatic Usage
+
+For custom providers or advanced setups, use the Python API directly:
+
+```python
+from agentskills_core import SkillRegistry
+from agentskills_mcp_server import create_mcp_server
+
+registry = SkillRegistry()
+await registry.register("incident-response", my_custom_provider)
+
+server = create_mcp_server(registry, name="My Skills Server")
+server.run()  # stdio by default
+```
+
+## Tools
+
+The server exposes tools that let the LLM agent access skill content:
+
+| Tool | Parameters | Description |
+| --- | --- | --- |
+| `get_skill_metadata` | `skill_id` | Read frontmatter (name, description, etc.) |
+| `get_skill_body` | `skill_id` | Load full skill instructions |
+| `get_skill_reference` | `skill_id`, `name` | Read a reference document |
+| `get_skill_script` | `skill_id`, `name` | Read a script |
+| `get_skill_asset` | `skill_id`, `name` | Read an asset |
+
+## Resources
+
+The server provides resources for system-prompt context:
+
+| URI | Description |
+| --- | --- |
+| `skills://catalog/xml` | XML catalog of all registered skills |
+| `skills://catalog/markdown` | Markdown catalog of all registered skills |
+| `skills://tools-usage-instructions` | Workflow instructions for using the tools |
+
+The MCP client reads these resources and injects them into the system prompt, giving the agent both *what* skills exist and *how* to interact with them.
+
+## API
+
+### `create_mcp_server(registry, *, name, instructions=None) -> FastMCP`
+
+| Parameter | Type | Description |
+| --- | --- | --- |
+| `registry` | `SkillRegistry` | The registry whose skills are exposed |
+| `name` | `str` | Display name for the MCP server (required) |
+| `instructions` | `str \| None` | Optional server-level instructions sent to clients |
+
+Returns a configured `FastMCP` instance ready for `server.run()`.
+
+## License
+
+MIT

agentskills_mcp_server-0.2.0/agentskills_mcp_server/__init__.py

@@ -0,0 +1,35 @@
+"""MCP server integration for Agent Skills.
+
+This package bridges :mod:`agentskills_core` and the `Model Context
+Protocol <https://modelcontextprotocol.io>`_, providing:
+
+* :func:`create_mcp_server` -- builds a FastMCP server from a
+  :class:`~agentskills_core.SkillRegistry`.  Useful when you have
+  custom providers or need full control over registration.
+* CLI entry-point (``python -m agentskills_mcp_server --config server.json``)
+  for zero-code server startup.
+
+Quick start (programmatic)::
+
+    from agentskills_core import SkillRegistry
+    from agentskills_mcp_server import create_mcp_server
+
+    registry = SkillRegistry()
+    await registry.register("incident-response", my_custom_provider)
+    server = create_mcp_server(registry, name="My Agent")
+    server.run()  # stdio by default
+
+CLI::
+
+    python -m agentskills_mcp_server --config server.json
+
+Install::
+
+    pip install agentskills-mcp-server
+"""
+
+from agentskills_mcp_server.server import create_mcp_server
+
+__all__ = [
+    "create_mcp_server",
+]

agentskills_mcp_server-0.2.0/agentskills_mcp_server/__main__.py

@@ -0,0 +1,124 @@
+"""Run an Agent Skills MCP server from a config file.
+
+Usage::
+
+    python -m agentskills_mcp_server --config server.json
+    python -m agentskills_mcp_server --config server.yaml
+    python -m agentskills_mcp_server --config server.json --transport streamable-http
+
+The config file is a JSON or YAML document conforming to
+:class:`~agentskills_mcp_server.config.ServerConfig`.
+
+Example ``server.json``::
+
+    {
+        "name": "My Skills Server",
+        "skills": [
+            {
+                "id": "incident-response",
+                "provider": "fs",
+                "options": {"root": "./skills"}
+            },
+            {
+                "id": "cloud-runbooks",
+                "provider": "http",
+                "options": {
+                    "base_url": "https://cdn.example.com/skills",
+                    "headers": {"Authorization": "Bearer <token>"},
+                    "params": {"sv": "2020-08-04", "sig": "<sas-token>"}
+                }
+            }
+        ]
+    }
+
+MCP client integration (stdio transport)::
+
+    {
+        "command": "python",
+        "args": ["-m", "agentskills_mcp_server", "--config", "server.json"]
+    }
+
+MCP client integration (streamable-http transport)::
+
+    {
+        "url": "http://127.0.0.1:8000/mcp"
+    }
+"""
+
+from __future__ import annotations
+
+import argparse
+import asyncio
+import json
+import sys
+from pathlib import Path
+
+
+def main() -> None:
+    """Parse CLI arguments, load config, and start the MCP server."""
+    parser = argparse.ArgumentParser(
+        prog="agentskills_mcp_server",
+        description="Start an Agent Skills MCP server from a config file.",
+    )
+    parser.add_argument(
+        "--config",
+        required=True,
+        type=Path,
+        help="Path to a JSON or YAML configuration file.",
+    )
+    parser.add_argument(
+        "--transport",
+        default="stdio",
+        choices=["stdio", "streamable-http"],
+        help="MCP transport type (default: stdio).",
+    )
+    args = parser.parse_args()
+
+    # ------------------------------------------------------------------
+    # Load config file
+    # ------------------------------------------------------------------
+    config_path: Path = args.config
+    if not config_path.exists():
+        print(
+            f"Error: config file not found: {config_path}",
+            file=sys.stderr,
+        )
+        sys.exit(1)
+
+    raw = config_path.read_text(encoding="utf-8")
+
+    if config_path.suffix in (".yaml", ".yml"):
+        try:
+            import yaml  # type: ignore[import-untyped]
+        except ImportError:
+            print(
+                "Error: YAML config files require pyyaml. Install with:  pip install pyyaml",
+                file=sys.stderr,
+            )
+            sys.exit(1)
+        data = yaml.safe_load(raw)
+    else:
+        data = json.loads(raw)
+
+    # ------------------------------------------------------------------
+    # Build and run
+    # ------------------------------------------------------------------
+    from agentskills_core import SkillRegistry
+    from agentskills_mcp_server.config import ServerConfig
+    from agentskills_mcp_server.server import _resolve_provider, create_mcp_server
+
+    config = ServerConfig(**data)
+
+    async def _build() -> object:
+        registry = SkillRegistry()
+        for skill_cfg in config.skills:
+            provider = _resolve_provider(skill_cfg.provider, skill_cfg.options)
+            await registry.register(skill_cfg.id, provider)
+        return create_mcp_server(registry, name=config.name, instructions=config.instructions)
+
+    server = asyncio.run(_build())
+    server.run(transport=args.transport)
+
+
+if __name__ == "__main__":
+    main()

agentskills_mcp_server-0.2.0/agentskills_mcp_server/config.py

@@ -0,0 +1,50 @@
+"""Pydantic configuration models for Agent Skills MCP servers.
+
+This module defines the declarative configuration schema used by
+the CLI (``python -m agentskills_mcp_server --config server.json``).
+
+Example config (JSON)::
+
+    {
+        "name": "My Skills Server",
+        "skills": [
+            {
+                "id": "incident-response",
+                "provider": "fs",
+                "options": {"root": "./skills"}
+            }
+        ]
+    }
+"""
+
+from __future__ import annotations
+
+from typing import Any
+
+from pydantic import BaseModel, Field
+
+
+class SkillConfig(BaseModel):
+    """Configuration for a single skill."""
+
+    id: str = Field(..., description="Skill identifier")
+    provider: str = Field(..., description="Provider type (e.g., 'fs', 'http')")
+    options: dict[str, Any] = Field(
+        default_factory=dict,
+        description="Provider-specific options passed to the provider constructor",
+    )
+
+
+class ServerConfig(BaseModel):
+    """Top-level configuration for an Agent Skills MCP server.
+
+    Attributes:
+        name: Display name shown to MCP clients during initialization.
+        instructions: Optional server-level instructions sent to the
+            client during the MCP handshake.
+        skills: One or more skill definitions to register.
+    """
+
+    name: str = Field(..., description="Display name for the MCP server")
+    instructions: str | None = Field(None, description="Optional server-level instructions")
+    skills: list[SkillConfig] = Field(..., description="Skills to register", min_length=1)

agentskills_mcp_server-0.2.0/agentskills_mcp_server/server.py

@@ -0,0 +1,248 @@
+"""MCP server builder for Agent Skills.
+
+This module creates a `FastMCP <https://pypi.org/project/mcp/>`_ server
+that exposes a :class:`~agentskills_core.SkillRegistry` as a set of MCP
+tools and resources.
+
+Tools
+-----
+Tools give the LLM agent access to skill content:
+
+==============================  =============================================
+Tool name                       Description
+==============================  =============================================
+``get_skill_metadata``          Read frontmatter (name, description, ...).
+``get_skill_body``              Load full skill instructions.
+``get_skill_reference``         Read a single reference document.
+``get_skill_script``            Read a single script.
+``get_skill_asset``             Read a single asset.
+==============================  =============================================
+
+Resources
+---------
+Resources provide context for the system prompt:
+
+==========================================  ==============================================
+URI                                         Description
+==========================================  ==============================================
+``skills://catalog/xml``                    XML catalog of all registered skills.
+``skills://catalog/markdown``               Markdown catalog of all registered skills.
+``skills://tools-usage-instructions``       Workflow instructions for using the tools.
+==========================================  ==============================================
+
+The developer reads these resources and injects them into the system
+prompt, giving the LLM agent both *what* skills exist and *how* to
+interact with them.
+
+Example::
+
+    from agentskills_mcp_server import create_mcp_server
+
+    server = create_mcp_server(registry, name="My Agent")
+    server.run()  # stdio by default
+"""
+
+from __future__ import annotations
+
+import json
+from pathlib import Path
+from typing import Any
+
+from mcp.server.fastmcp import FastMCP
+
+from agentskills_core import SkillProvider, SkillRegistry
+
+# ------------------------------------------------------------------
+# Provider resolution
+# ------------------------------------------------------------------
+
+#: Provider types that are recognized by :func:`_resolve_provider`.
+SUPPORTED_PROVIDERS: frozenset[str] = frozenset({"fs", "http"})
+
+
+def _resolve_provider(provider_type: str, options: dict[str, Any]) -> SkillProvider:
+    """Map a provider type string and options to a concrete provider.
+
+    Args:
+        provider_type: One of the :data:`SUPPORTED_PROVIDERS` keys.
+        options: Keyword arguments forwarded to the provider
+            constructor.  Unknown keys are silently ignored for
+            safety (e.g. a ``client`` key cannot be serialized
+            to JSON and must not be passed).
+
+    Returns:
+        A ready-to-use :class:`~agentskills_core.SkillProvider`.
+
+    Raises:
+        ImportError: If the required provider package is not installed.
+        ValueError: If *provider_type* is not recognized.
+    """
+    if provider_type == "fs":
+        try:
+            from agentskills_fs import LocalFileSystemSkillProvider
+        except ImportError as exc:
+            raise ImportError(
+                "Provider 'fs' requires the agentskills-fs package. "
+                "Install it with:  pip install agentskills-fs"
+            ) from exc
+        root = Path(options.get("root", "."))
+        return LocalFileSystemSkillProvider(root=root)
+
+    if provider_type == "http":
+        try:
+            from agentskills_http import HTTPStaticFileSkillProvider
+        except ImportError as exc:
+            raise ImportError(
+                "Provider 'http' requires the agentskills-http package. "
+                "Install it with:  pip install agentskills-http"
+            ) from exc
+        # Only pass constructor-safe keys; runtime objects like
+        # ``client`` cannot be serialized to a config file.
+        safe_http_keys = {"base_url", "headers", "params"}
+        filtered = {k: v for k, v in options.items() if k in safe_http_keys}
+        return HTTPStaticFileSkillProvider(**filtered)
+
+    raise ValueError(
+        f"Unknown provider type: {provider_type!r}. "
+        f"Supported types: {', '.join(sorted(SUPPORTED_PROVIDERS))}"
+    )
+
+
+# ------------------------------------------------------------------
+# Server builder
+# ------------------------------------------------------------------
+
+
+def create_mcp_server(
+    registry: SkillRegistry,
+    *,
+    name: str,
+    instructions: str | None = None,
+) -> FastMCP:
+    """Build an MCP server that exposes an Agent Skills registry.
+
+    The returned :class:`~mcp.server.fastmcp.FastMCP` server is
+    transport-agnostic.  Call ``server.run()`` to start with the
+    default stdio transport, or ``server.run(transport="streamable-http")``
+    for HTTP.
+
+    **Tools** let the LLM agent read skill content (metadata,
+    body, references, scripts, assets).  **Resources** provide
+    the skill catalog and usage instructions for system-prompt
+    injection.
+
+    Args:
+        registry: The :class:`~agentskills_core.SkillRegistry` whose
+            skills should be exposed via MCP.
+        name: Display name for the MCP server.  Clients see this
+            during the MCP initialization handshake.  Required.
+        instructions: Optional server-level instructions sent to the
+            MCP client during initialization.  Use this to describe
+            the server's purpose or capabilities.
+
+    Returns:
+        A configured :class:`~mcp.server.fastmcp.FastMCP` server
+        instance, ready for ``server.run()``.
+    """
+    mcp = FastMCP(name, instructions=instructions)
+
+    # ------------------------------------------------------------------
+    # Tools
+    # ------------------------------------------------------------------
+
+    @mcp.tool()
+    async def get_skill_metadata(skill_id: str) -> str:
+        """Get structured metadata (name, description, and optional fields like license, compatibility, metadata) for a specific skill."""  # noqa: E501
+        skill = registry.get_skill(skill_id)
+        return json.dumps(await skill.get_metadata())
+
+    @mcp.tool()
+    async def get_skill_body(skill_id: str) -> str:
+        """Get the full instructions and guidance (markdown body) for a specific skill."""
+        skill = registry.get_skill(skill_id)
+        return await skill.get_body()
+
+    @mcp.tool()
+    async def get_skill_reference(skill_id: str, name: str) -> str:
+        """Get the full content of a specific reference document from a skill.
+
+        Provide both skill_id and the reference name.  Binary content is
+        decoded as UTF-8 with replacement characters for non-decodable bytes.
+        """
+        skill = registry.get_skill(skill_id)
+        return (await skill.get_reference(name)).decode("utf-8", errors="replace")
+
+    @mcp.tool()
+    async def get_skill_asset(skill_id: str, name: str) -> str:
+        """Get the content of a specific asset from a skill.
+
+        Provide both skill_id and the asset name.  Binary content is
+        decoded as UTF-8 with replacement characters for non-decodable bytes.
+        """
+        skill = registry.get_skill(skill_id)
+        return (await skill.get_asset(name)).decode("utf-8", errors="replace")
+
+    @mcp.tool()
+    async def get_skill_script(skill_id: str, name: str) -> str:
+        """Get the content of a specific script from a skill.
+
+        Provide both skill_id and the script name.  Binary content is
+        decoded as UTF-8 with replacement characters for non-decodable bytes.
+        """
+        skill = registry.get_skill(skill_id)
+        return (await skill.get_script(name)).decode("utf-8", errors="replace")
+
+    # ------------------------------------------------------------------
+    # Resources
+    # ------------------------------------------------------------------
+
+    @mcp.resource("skills://catalog/xml")
+    async def skills_catalog_xml() -> str:
+        """XML catalog of all registered skills for system-prompt injection."""
+        return await registry.get_skills_catalog(format="xml")
+
+    @mcp.resource("skills://catalog/markdown")
+    async def skills_catalog_markdown() -> str:
+        """Markdown catalog of all registered skills for system-prompt injection."""
+        return await registry.get_skills_catalog(format="markdown")
+
+    @mcp.resource("skills://tools-usage-instructions")
+    def skills_tools_usage_instructions() -> str:
+        """Workflow instructions explaining how to use the Agent Skills tools."""
+        return _TOOLS_USAGE_INSTRUCTIONS
+
+    return mcp
+
+
+_TOOLS_USAGE_INSTRUCTIONS = """\
+## How to Use Agent Skills
+
+You have access to a set of **Agent Skills** — curated knowledge \
+bundles that contain step-by-step instructions, reference documents, \
+scripts, and assets. The available skills are listed in the catalog.
+
+### Workflow
+
+1. **Pick a skill** — Choose the most relevant skill from the catalog \
+based on the user's request.
+2. **Read metadata** — Call `get_skill_metadata(skill_id)` to get \
+structured information (name, description, and optional fields).
+3. **Read the body** — Call `get_skill_body(skill_id)` to load the \
+full instructions. Follow these instructions carefully.
+4. **Fetch resources on demand** — The skill body will reference \
+specific resources by name. Use the appropriate tool to retrieve them:
+   - `get_skill_reference(skill_id, name)` — reference documents \
+(policies, templates, runbooks)
+   - `get_skill_script(skill_id, name)` — executable scripts
+   - `get_skill_asset(skill_id, name)` — diagrams, data files, or \
+other assets
+
+### Important guidelines
+
+- **Do not guess resource names.** Only fetch resources that are \
+explicitly mentioned in the skill body.
+- **Follow progressive disclosure.** Read the skill body first, then \
+fetch only the resources you need for the current step.
+- **One skill at a time.** Focus on the most relevant skill for the \
+user's request. If multiple skills apply, address them sequentially.\
+"""

agentskills_mcp_server-0.2.0/pyproject.toml

@@ -0,0 +1,34 @@
+[tool.poetry]
+name = "agentskills-mcp-server"
+version = "0.2.0"
+description = "MCP server integration for the Agent Skills format — expose skills as MCP tools and resources (https://agentskills.io)"
+license = "MIT"
+authors = ["Pratik Panda"]
+readme = "README.md"
+homepage = "https://agentskills.io"
+repository = "https://github.com/pratikxpanda/agentskills-sdk"
+packages = [{include = "agentskills_mcp_server"}]
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Intended Audience :: Developers",
+    "Topic :: Software Development :: Libraries",
+    "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13",
+]
+
+[tool.poetry.dependencies]
+python = ">=3.12,<4.0"
+agentskills-core = ">=0.1.0"
+mcp = ">=1.0"
+pydantic = ">=2.0"
+
+[tool.poetry.extras]
+fs = ["agentskills-fs"]
+http = ["agentskills-http"]
+
+[tool.poetry.scripts]
+agentskills-mcp-server = "agentskills_mcp_server.__main__:main"
+
+[build-system]
+requires = ["poetry-core"]
+build-backend = "poetry.core.masonry.api"