Add Skills extension

docs/advanced/skills.md

@@ -0,0 +1,132 @@
+# Skills
+
+[SEP-2640](https://github.com/modelcontextprotocol/modelcontextprotocol/pull/2640) defines a
+convention for serving [Agent Skills](https://agentskills.io/) over MCP. A skill is just a
+directory of files — at minimum a `SKILL.md` with YAML frontmatter — that you expose as ordinary
+MCP resources, conventionally under a `skill://` URI.
+
+A server enumerates its skills with `skills/list`, answers for any single one by URI with
+`skills/get`, and — optionally — lists a directory's direct children with
+`resources/directory/read`.
+
+The SDK ships this as the built-in `Skills` extension (`io.modelcontextprotocol/skills`). There's
+one on the server side and one on the client side. If [Extensions](extensions.md) are new to you,
+skim that page first.
+
+!!! info
+    `Skills` gives you the **protocol** primitives: request/response handling, capability
+    advertisement, and SEP-2640 conformance validation.
+
+    It does **not** discover, read, or hash skills from a filesystem. You supply handlers that
+    answer from wherever your catalog actually lives — a database, a generated index, an in-memory
+    list, or a directory you walk yourself — and you serve each skill's files as ordinary resources
+    through `MCPServer.add_resource` or an `@mcp.resource(...)` template handler.
+
+## Serving a skill
+
+Here's a server that serves one skill:
+
+```python title="server.py" hl_lines="31-41 44-51 55"
+--8<-- "docs_src/skills/tutorial001.py"
+```
+
+There are three moves here:
+
+* `Skill(uri=..., frontmatter=..., resources=[...])` is one entry. It has the same shape whether it
+  comes back from `skills/list` or `skills/get`. `resources` is the skill's complete file
+  manifest — every file, `SKILL.md` included, each with a `sha256:...` digest and byte size — or
+  the string `"dynamic"` for content generated on demand.
+* `list_skills` and `get_skill` are plain async callables, invoked once per request. `get_skill`
+  **must** answer for a skill even if your `list_skills` left it out — SEP-2640 requires a server
+  to answer by URI for every skill it serves, listed or not.
+* `mcp.add_resource(TextResource(uri=SKILL_URI, ...))` registers the skill's actual file content,
+  served through the SDK's ordinary resource machinery. `Skills` never reads or writes resource
+  content itself.
+
+And that's it. `Skills(list_skills=..., get_skill=...)` is all a server needs;
+`resources/directory/read` is optional (more on that below).
+
+## Fetching a skill
+
+On the client side, `Skills` is a [`ClientExtension`](extensions.md). You register it the same way
+you register any other one — by passing it to `Client(extensions=[...])` — and then call `bind` to
+get the verbs tied to that connection:
+
+```python title="client.py" hl_lines="9-11"
+--8<-- "docs_src/skills/tutorial001_client.py"
+```
+
+`skills.bind(client)` hands you a `BoundSkills`, and its methods are the SEP-2640 verbs:
+
+* `list_skills` and `read_directory` follow `nextCursor` to completion, so a single call gives you
+  every page's skills or resources.
+* `get_skill` costs exactly one request.
+
+These three validate the server's response against the SEP-2640 conformance rules before returning
+it. A name that doesn't match its URI, a digest in the wrong shape, or an incomplete manifest raises
+`ValueError` rather than reaching your code.
+
+`read_skill_uri` is the exception. It's a thin, discoverable alias for `resources/read` that returns
+a `ReadResourceResult` (text or blob contents) and validates nothing itself — that's the next step.
+
+!!! tip
+    `verify_skill_resource(skill, uri, content)` checks a file's bytes — size, then SHA-256
+    digest — against the entry you hold for it. Call it after `read_skill_uri` and *before* you
+    treat the content as trustworthy.
+
+    `resources/read` returns whatever bytes the server sends *right now*; verification is what ties
+    those bytes back to the manifest you already validated. It applies to a static manifest only —
+    a `"dynamic"` skill carries no digests, so calling it on one raises `ValueError`.
+
+!!! warning
+    Skill content is untrusted model input, exactly like any other server-provided text. SEP-2640
+    requires a host to tag it with its originating server before it reaches the model, and to
+    never grant the frontmatter's `allowed-tools` field (or any other permission-widening field)
+    without explicit per-skill user approval.
+
+    Both are host responsibilities the SDK cannot discharge for you — read the SEP's
+    [Security Implications](https://github.com/modelcontextprotocol/modelcontextprotocol/pull/2640)
+    section before building a host on top of this extension.
+
+## Directory reads
+
+A skill's instructions often point at a directory rather than a file ("pick the matching
+template from `templates/`"). `resources/list` can't answer that — it enumerates a server's
+entire resource space, not one subtree — so SEP-2640 adds `resources/directory/read`, gated
+behind the `directoryRead` capability setting:
+
+```python
+mcp = MCPServer(
+    "catalog",
+    extensions=[
+        Skills(
+            list_skills=list_skills,
+            get_skill=get_skill,
+            read_directory=read_directory,  # lists uri's direct children
+        )
+    ],
+)
+```
+
+Supplying `read_directory` advertises `{"directoryRead": true}` under the extension's
+capabilities. Omitting it advertises neither the setting nor the method — a client calling
+`resources/directory/read` against such a server gets `METHOD_NOT_FOUND`. On the client side,
+`read_directory` raises before it sends anything if the connected server hasn't advertised the
+setting.
+
+## Protocol version and caching
+
+In protocol version `2026-07-28` and later, `skills/list` and `skills/get` results carry the base
+protocol's caching fields, [`ttlMs` and `cacheScope`](https://github.com/modelcontextprotocol/modelcontextprotocol/pull/2549) —
+the same freshness hint `tools/list`, `resources/list`, and `resources/read` carry. `Skills` fills
+`cacheScope` with `"public"` when your handler leaves it unset, and omits both fields entirely on an
+older connection. So you don't have to branch on protocol version yourself.
+
+## What this SDK doesn't do
+
+`Skills` is a protocol adapter, not a skills provider. It has no opinion on where a skill's bytes
+live, how they're indexed, or when a catalog is refreshed — that's for a higher-level library, or
+your own handler, to decide.
+
+If you're looking for "scan this directory and serve whatever's in it," you're looking for a
+provider built on top of `Skills`, not `Skills` itself.

docs_src/skills/tutorial001.py

@@ -0,0 +1,54 @@
+import hashlib
+from typing import Any
+
+from mcp.server.context import ServerRequestContext
+from mcp.server.mcpserver import MCPServer
+from mcp.server.mcpserver.resources import TextResource
+from mcp.server.skills import Skills
+from mcp.shared.exceptions import MCPError
+from mcp.shared.skills import (
+    GetSkillParams,
+    GetSkillResult,
+    ListSkillsParams,
+    ListSkillsResult,
+    Skill,
+    SkillResource,
+)
+from mcp.types import INVALID_PARAMS
+
+SKILL_URI = "skill://git-workflow/SKILL.md"
+SKILL_MD = """\
+---
+name: git-workflow
+description: Follow this team's Git conventions for branching and commits
+---
+
+Branch from `main` using `type/short-description`. Write commit subjects in the
+imperative mood, under 72 characters.
+"""
+
+GIT_WORKFLOW = Skill(
+    uri=SKILL_URI,
+    frontmatter={"name": "git-workflow", "description": "Follow this team's Git conventions for branching and commits"},
+    resources=[
+        SkillResource(
+            uri=SKILL_URI,
+            digest=f"sha256:{hashlib.sha256(SKILL_MD.encode()).hexdigest()}",
+            size=len(SKILL_MD.encode()),
+        )
+    ],
+)
+
+
+async def list_skills(ctx: ServerRequestContext[Any, Any], params: ListSkillsParams) -> ListSkillsResult:
+    return ListSkillsResult(skills=[GIT_WORKFLOW])
+
+
+async def get_skill(ctx: ServerRequestContext[Any, Any], params: GetSkillParams) -> GetSkillResult:
+    if params.uri != SKILL_URI:
+        raise MCPError(code=INVALID_PARAMS, message=f"unknown skill: {params.uri}")
+    return GetSkillResult(skill=GIT_WORKFLOW)
+
+
+mcp = MCPServer("catalog", extensions=[Skills(list_skills=list_skills, get_skill=get_skill)])
+mcp.add_resource(TextResource(uri=SKILL_URI, name="SKILL.md", mime_type="text/markdown", text=SKILL_MD))

docs_src/skills/tutorial001_client.py

@@ -0,0 +1,24 @@
+import anyio
+
+from mcp import Client
+from mcp.client.skills import Skills, verify_skill_resource
+from mcp.types import TextResourceContents
+
+
+async def main() -> None:
+    skills = Skills()
+    async with Client("http://localhost:8000/mcp", extensions=[skills]) as client:
+        catalog = skills.bind(client)
+        for skill in await catalog.list_skills():
+            print(skill.uri, skill.frontmatter["description"])
+
+        skill = await catalog.get_skill("skill://git-workflow/SKILL.md")
+        result = await catalog.read_skill_uri(skill.uri)
+        content = result.contents[0]
+        if isinstance(content, TextResourceContents):
+            verify_skill_resource(skill, skill.uri, content.text.encode())
+            print(content.text)
+
+
+if __name__ == "__main__":
+    anyio.run(main)

mkdocs.yml

@@ -67,6 +67,7 @@ nav:
       - Middleware: advanced/middleware.md
       - Extensions: advanced/extensions.md
       - MCP Apps: advanced/apps.md
+      - Skills: advanced/skills.md
   - Troubleshooting: troubleshooting.md
   - Translations: translations.md
   - Migration Guide: migration.md