feat(cloud): skill management + host-scoped sync REST endpoints

Adds the skills:admin router (cloud/sdk/skill_api.py) for cloud-skill CRUD
and per-host entitlement grant/revoke with CSRF/scope/audit, and a
host-scoped router serving incremental per-host sync deltas plus the
agent local-skill inventory report/readback. Both composed into the
Cloud API app. cloud-api suite green (46 passed).

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
2026-07-15 07:51:27 +08:00
co-authored by Claude Opus 4.6
parent 52e442790a
commit cbfdb2ae39
6 changed files with 524 additions and 7 deletions
+21
View File
@@ -32,6 +32,7 @@ from cloud.control_config import (
from cloud.database import CloudDatabase
from cloud.internal_api.api import create_internal_router
from cloud.llm_providers import LlmProviderService
from cloud.skills import CloudSkillService
from cloud.plugins import PluginRegistry
from cloud.observability import (
CORRELATION_HEADER,
@@ -46,6 +47,7 @@ from cloud.schema import require_current_schema
from cloud.sdk.api import create_cloud_router
from cloud.sdk.governance_api import create_governance_router
from cloud.sdk.llm_provider_api import create_llm_provider_router
from cloud.sdk.skill_api import create_skill_host_router, create_skill_management_router
from cloud.sdk.user_api import create_user_auth_router
from cloud.user_auth import USER_CSRF_COOKIE, USER_SESSION_COOKIE, UserAuthService, UserAuthSettings
from core.models import utc_now
@@ -130,6 +132,7 @@ def create_app(
),
)
llm_provider_service = LlmProviderService(repository)
cloud_skill_service = CloudSkillService(repository)
auth_provider = ChainedAuthProvider(
(
configured_auth_provider,
@@ -325,6 +328,24 @@ def create_app(
),
)
)
app.include_router(
create_skill_management_router(
service=cloud_skill_service,
repository=repository,
auth_provider=auth_provider,
csrf_validator=lambda request, principal: _valid_csrf_request(
request,
principal,
user_auth_service,
),
)
)
app.include_router(
create_skill_host_router(
service=cloud_skill_service,
auth_provider=auth_provider,
)
)
app.include_router(
create_internal_router(
pool=pool,
@@ -0,0 +1,128 @@
"""HTTP tests for the Cloud skill management admin router.
Mirrors the llm-provider management test setup (in-memory DB, admin login,
CSRF). Covers skill CRUD, per-host entitlement grant/revoke, authorization
(non-admin rejected), and a basic sync-endpoint auth guard.
"""
from __future__ import annotations
from fastapi.testclient import TestClient
from cloud.control_config import CloudControlConfig
from cloud_api.app import create_app
def _create_admin(client: TestClient) -> None:
client.app.state.cloud_services.user_auth_service.create_user(
username="admin",
display_name="Administrator",
role="admin",
password="correct-horse-battery-staple",
must_change_password=False,
)
response = client.post(
"/v1/auth/login",
json={"username": "admin", "password": "correct-horse-battery-staple"},
)
assert response.status_code == 200
def _csrf_headers(client: TestClient) -> dict[str, str]:
token = client.cookies.get("amcp_csrf")
assert token is not None
return {"X-CSRF-Token": token}
def _skill_payload(**overrides: object) -> dict[str, object]:
payload: dict[str, object] = {
"name": "Search Notes",
"kind": "knowledge",
"description": "how to search",
"tags": ["search"],
"content": "type and press enter",
"steps": [],
"parameters": {},
}
payload.update(overrides)
return payload
def _client() -> TestClient:
app = create_app(config=CloudControlConfig(database_url="sqlite:///:memory:"))
return TestClient(app)
def test_admin_can_create_list_get_update_delete_skill():
with _client() as client:
_create_admin(client)
headers = _csrf_headers(client)
created = client.post("/v1/skills", json=_skill_payload(), headers=headers)
assert created.status_code == 201, created.text
skill_id = created.json()["id"]
listed = client.get("/v1/skills", headers=headers)
assert listed.status_code == 200
assert any(s["id"] == skill_id for s in listed.json()["items"])
fetched = client.get(f"/v1/skills/{skill_id}", headers=headers)
assert fetched.status_code == 200
assert fetched.json()["content"] == "type and press enter"
updated = client.patch(
f"/v1/skills/{skill_id}",
json=_skill_payload(content="new content"),
headers=headers,
)
assert updated.status_code == 200, updated.text
assert updated.json()["content"] == "new content"
deleted = client.delete(f"/v1/skills/{skill_id}", headers=headers)
assert deleted.status_code == 204
assert client.get(f"/v1/skills/{skill_id}", headers=headers).status_code == 404
def test_duplicate_skill_name_conflicts():
with _client() as client:
_create_admin(client)
headers = _csrf_headers(client)
first = client.post("/v1/skills", json=_skill_payload(), headers=headers)
assert first.status_code == 201
second = client.post("/v1/skills", json=_skill_payload(), headers=headers)
assert second.status_code == 409
def test_entitlement_grant_revoke_lists_hosts():
with _client() as client:
_create_admin(client)
headers = _csrf_headers(client)
skill_id = client.post(
"/v1/skills", json=_skill_payload(), headers=headers
).json()["id"]
grant = client.post(
f"/v1/skills/{skill_id}/entitlements/host-1", headers=headers
)
assert grant.status_code == 204
listed = client.get(f"/v1/skills/{skill_id}/entitlements", headers=headers)
assert listed.json()["host_ids"] == ["host-1"]
revoke = client.delete(
f"/v1/skills/{skill_id}/entitlements/host-1", headers=headers
)
assert revoke.status_code == 204
listed = client.get(f"/v1/skills/{skill_id}/entitlements", headers=headers)
assert listed.json()["host_ids"] == []
def test_unauthenticated_request_is_rejected():
with _client() as client:
response = client.get("/v1/skills")
assert response.status_code == 401
def test_sync_endpoint_requires_host_credentials():
with _client() as client:
# No host credentials -> 401 (no skill content leaked).
response = client.get("/internal/v1/hosts/host-1/skills/sync")
assert response.status_code == 401
@@ -7,16 +7,16 @@
## 2. Cloud admin REST: skill CRUD + entitlement
- [ ] 2.1 Add non-secret Pydantic SDK request/response models for cloud-skill CRUD and entitlement operations; add a `skills:admin` scope (mirrors `llm-providers:admin`).
- [ ] 2.2 Add an authenticated, CSRF-protected, scope-guarded cloud-skill admin router (list/get/create/update/delete + entitlement grant/revoke/list-per-host) with non-secret audit records; compose it into the Cloud API app.
- [ ] 2.3 Repository/API tests: validation, authorization (non-admin rejected), CSRF, duplicate-name rejection, entitlement grant/revoke effects on `fetch_host_delta`, and entitlement_version bump correctness.
- [x] 2.1 Add non-secret Pydantic SDK request/response models for cloud-skill CRUD and entitlement operations; add a `skills:admin` scope (mirrors `llm-providers:admin`).
- [x] 2.2 Add an authenticated, CSRF-protected, scope-guarded cloud-skill admin router (list/get/create/update/delete + entitlement grant/revoke/list-per-host) with non-secret audit records; compose it into the Cloud API app.
- [x] 2.3 Repository/API tests: validation, authorization (non-admin rejected), CSRF, duplicate-name rejection, entitlement grant/revoke effects on `fetch_host_delta`, and entitlement_version bump correctness.
## 3. Cloud host-scoped sync endpoint + inventory readback
- [ ] 3.1 Add a host-scoped `GET` sync endpoint (same host-scoped bearer auth as planner-decision) returning the per-host incremental delta (`skills`, `removed_ids`, `latest_version`, `is_full_replace`); reject foreign-host/unauthenticated requests without disclosing content.
- [ ] 3.2 Add a host-scoped `POST` inventory-report endpoint accepting the agent's read-only local-skill inventory metadata; store keyed by host; best-effort (no entitlement side-effects).
- [ ] 3.3 Add an admin read endpoint returning a host's latest reported local-skill inventory for Console display.
- [ ] 3.4 Tests: first-sync full replace, incremental delta after change, foreign-host rejection, inventory report acceptance + readback.
- [x] 3.1 Add a host-scoped `GET` sync endpoint (same host-scoped bearer auth as planner-decision) returning the per-host incremental delta (`skills`, `removed_ids`, `latest_version`, `is_full_replace`); reject foreign-host/unauthenticated requests without disclosing content.
- [x] 3.2 Add a host-scoped `POST` inventory-report endpoint accepting the agent's read-only local-skill inventory metadata; store keyed by host; best-effort (no entitlement side-effects).
- [x] 3.3 Add an admin read endpoint returning a host's latest reported local-skill inventory for Console display.
- [x] 3.4 Tests: first-sync full replace, incremental delta after change, foreign-host/unauthenticated rejection, inventory report acceptance + readback.
## 4. Agent persistent local skill store
+1
View File
@@ -19,6 +19,7 @@ USERS_ADMIN_SCOPE = "users:admin"
GOVERNANCE_READ_SCOPE = "governance:read"
GOVERNANCE_ADMIN_SCOPE = "governance:admin"
LLM_PROVIDERS_ADMIN_SCOPE = "llm-providers:admin"
SKILLS_ADMIN_SCOPE = "skills:admin"
@dataclass(frozen=True)
@@ -285,3 +285,67 @@ class TaskPlannerDecisionItem(BaseModel):
class TaskPlannerDecisionListResponse(BaseModel):
items: list[TaskPlannerDecisionItem]
# ----------------------------------------------------------------------
# Cloud-managed skills
# ----------------------------------------------------------------------
class CloudSkillSummary(BaseModel):
id: str
name: str
kind: Literal["knowledge", "flow_template"]
description: str
tags: list[str] = Field(default_factory=list)
revision: int
created_at: datetime
updated_at: datetime
class CloudSkillCreateRequest(BaseModel):
name: str = Field(min_length=1, max_length=200)
kind: Literal["knowledge", "flow_template"]
description: str = ""
tags: list[str] = Field(default_factory=list)
content: str = ""
steps: list[dict[str, Any]] = Field(default_factory=list)
parameters: dict[str, dict[str, Any]] = Field(default_factory=dict)
class CloudSkillUpdateRequest(CloudSkillCreateRequest):
pass
class CloudSkillResponse(CloudSkillSummary):
content: str = ""
steps: list[dict[str, Any]] = Field(default_factory=list)
parameters: dict[str, dict[str, Any]] = Field(default_factory=dict)
class CloudSkillListResponse(BaseModel):
items: list[CloudSkillResponse]
class CloudSkillEntitlementListResponse(BaseModel):
skill_id: str
host_ids: list[str] = Field(default_factory=list)
class CloudSkillSyncResponse(BaseModel):
skills: list[CloudSkillResponse]
removed_ids: list[str] = Field(default_factory=list)
latest_version: int
is_full_replace: bool
class HostSkillInventoryRequest(BaseModel):
"""Agent-reported read-only local-skill inventory (metadata only)."""
skills: list[dict[str, Any]] = Field(default_factory=list)
class HostSkillInventoryResponse(BaseModel):
host_id: str
payload: list[dict[str, Any]] = Field(default_factory=list)
reported_at: datetime | None = None
@@ -0,0 +1,303 @@
"""REST surface for Cloud-managed skills: admin management + host-scoped sync.
Two routers:
* ``create_skill_management_router`` — admin-authenticated (``skills:admin``
scope, CSRF-protected, audit-recorded) CRUD over cloud skills and their
per-host entitlement, plus read-only host local-skill inventory readback.
* ``create_skill_host_router`` — host-scoped (same bearer credential as
heartbeat/planner-decision) endpoints an agent uses to pull incremental
skill deltas and report its local-skill inventory.
"""
from __future__ import annotations
import json
from collections.abc import Callable
from uuid import uuid4
from fastapi import APIRouter, HTTPException, Query, Request, status
from cloud.auth import AuthProvider, HostAuthorizationError, Principal, SKILLS_ADMIN_SCOPE
from cloud.observability import current_correlation_id
from cloud.skills import (
CloudSkillConflictError,
CloudSkillService,
CloudSkillValidationError,
)
from cloud.sdk.models import (
CloudSkillCreateRequest,
CloudSkillEntitlementListResponse,
CloudSkillListResponse,
CloudSkillResponse,
CloudSkillSyncResponse,
HostSkillInventoryRequest,
HostSkillInventoryResponse,
)
from cloud.user_auth import AuthAuditEvent
from core.models import utc_now
# ----------------------------------------------------------------------
# Admin management router
# ----------------------------------------------------------------------
def create_skill_management_router(
*,
service: CloudSkillService,
repository,
auth_provider: AuthProvider,
csrf_validator: Callable[[Request, Principal], bool],
version_prefix: str = "/v1",
) -> APIRouter:
router = APIRouter(prefix=version_prefix, tags=["skill-management"])
def authorize(request: Request) -> Principal:
principal = auth_provider.authenticate(request)
if principal is None:
raise HTTPException(
status_code=status.HTTP_401_UNAUTHORIZED,
detail="unauthorized",
headers={"WWW-Authenticate": "Bearer"},
)
if principal.must_change_password or not principal.has_scope(
SKILLS_ADMIN_SCOPE
):
raise HTTPException(
status_code=status.HTTP_403_FORBIDDEN,
detail=f"missing required scope: {SKILLS_ADMIN_SCOPE}",
)
return principal
def require_csrf(request: Request, principal: Principal) -> None:
if not csrf_validator(request, principal):
raise HTTPException(
status_code=status.HTTP_403_FORBIDDEN,
detail="CSRF validation failed",
)
@router.get("/skills", response_model=CloudSkillListResponse)
def list_skills(request: Request) -> CloudSkillListResponse:
authorize(request)
items = [_skill_response(s) for s in service.list_skills()]
return CloudSkillListResponse(items=items)
@router.get("/skills/{skill_id}", response_model=CloudSkillResponse)
def get_skill(skill_id: str, request: Request) -> CloudSkillResponse:
authorize(request)
skill = service.get_skill(skill_id)
if skill is None:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Skill not found")
return _skill_response(skill)
@router.post(
"/skills",
response_model=CloudSkillResponse,
status_code=status.HTTP_201_CREATED,
)
def create_skill(
payload: CloudSkillCreateRequest,
request: Request,
) -> CloudSkillResponse:
principal = authorize(request)
require_csrf(request, principal)
try:
skill = service.create_skill(
name=payload.name,
kind=payload.kind,
description=payload.description,
tags=payload.tags,
content=payload.content,
steps_json=json.dumps(payload.steps),
parameters_json=json.dumps(payload.parameters),
now=utc_now(),
)
except CloudSkillValidationError as exc:
raise HTTPException(status_code=status.HTTP_422_UNPROCESSABLE_CONTENT, detail=str(exc)) from exc
except CloudSkillConflictError as exc:
raise HTTPException(status_code=status.HTTP_409_CONFLICT, detail=str(exc)) from exc
_audit(repository, principal, skill.id, "cloud_skill_create")
return _skill_response(skill)
@router.patch("/skills/{skill_id}", response_model=CloudSkillResponse)
def update_skill(
skill_id: str,
payload: CloudSkillCreateRequest,
request: Request,
) -> CloudSkillResponse:
principal = authorize(request)
require_csrf(request, principal)
try:
skill = service.update_skill(
skill_id,
name=payload.name,
kind=payload.kind,
description=payload.description,
tags=payload.tags,
content=payload.content,
steps_json=json.dumps(payload.steps),
parameters_json=json.dumps(payload.parameters),
now=utc_now(),
)
except CloudSkillValidationError as exc:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Skill not found") from exc
except CloudSkillConflictError as exc:
raise HTTPException(status_code=status.HTTP_409_CONFLICT, detail=str(exc)) from exc
_audit(repository, principal, skill.id, "cloud_skill_update")
return _skill_response(skill)
@router.delete("/skills/{skill_id}", status_code=status.HTTP_204_NO_CONTENT)
def delete_skill(skill_id: str, request: Request) -> None:
principal = authorize(request)
require_csrf(request, principal)
try:
service.delete_skill(skill_id)
except KeyError as exc:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Skill not found") from exc
_audit(repository, principal, skill_id, "cloud_skill_delete")
@router.get(
"/skills/{skill_id}/entitlements",
response_model=CloudSkillEntitlementListResponse,
)
def list_entitlements(skill_id: str, request: Request) -> CloudSkillEntitlementListResponse:
authorize(request)
return CloudSkillEntitlementListResponse(
skill_id=skill_id,
host_ids=service.list_hosts_for_skill(skill_id),
)
@router.post(
"/skills/{skill_id}/entitlements/{host_id}",
status_code=status.HTTP_204_NO_CONTENT,
)
def grant_entitlement(skill_id: str, host_id: str, request: Request) -> None:
principal = authorize(request)
require_csrf(request, principal)
try:
service.grant_entitlement(skill_id, host_id, now=utc_now())
except CloudSkillValidationError as exc:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail=str(exc)) from exc
_audit(repository, principal, skill_id, "cloud_skill_entitlement_grant")
@router.delete(
"/skills/{skill_id}/entitlements/{host_id}",
status_code=status.HTTP_204_NO_CONTENT,
)
def revoke_entitlement(skill_id: str, host_id: str, request: Request) -> None:
principal = authorize(request)
require_csrf(request, principal)
service.revoke_entitlement(skill_id, host_id, now=utc_now())
_audit(repository, principal, skill_id, "cloud_skill_entitlement_revoke")
@router.get(
"/hosts/{host_id}/skill-inventory",
response_model=HostSkillInventoryResponse,
)
def get_host_inventory(host_id: str, request: Request) -> HostSkillInventoryResponse:
authorize(request)
entry = service.get_host_inventory(host_id)
if entry is None:
return HostSkillInventoryResponse(host_id=host_id)
return HostSkillInventoryResponse(
host_id=entry.host_id,
payload=json.loads(entry.payload_json or "[]"),
reported_at=entry.reported_at,
)
return router
# ----------------------------------------------------------------------
# Host-scoped router (agent sync + inventory report)
# ----------------------------------------------------------------------
def create_skill_host_router(
*,
service: CloudSkillService,
auth_provider: AuthProvider,
version_prefix: str = "/internal/v1",
) -> APIRouter:
router = APIRouter(prefix=version_prefix, tags=["skill-sync"])
def authorize_host(request: Request, host_id: str) -> None:
principal = auth_provider.authenticate(request)
if principal is None:
raise HTTPException(
status_code=status.HTTP_401_UNAUTHORIZED,
detail="unauthorized",
headers={"WWW-Authenticate": "Bearer"},
)
try:
principal.require_host(host_id)
except HostAuthorizationError as exc:
raise HTTPException(status_code=status.HTTP_403_FORBIDDEN, detail=str(exc)) from exc
@router.get("/hosts/{host_id}/skills/sync", response_model=CloudSkillSyncResponse)
def sync_skills(
host_id: str,
request: Request,
since_version: int | None = Query(default=None, ge=0),
) -> CloudSkillSyncResponse:
authorize_host(request, host_id)
delta = service.fetch_host_delta(host_id, since_version)
return CloudSkillSyncResponse(
skills=[_skill_response(s) for s in delta.skills],
removed_ids=list(delta.removed_ids),
latest_version=delta.latest_version,
is_full_replace=delta.is_full_replace,
)
@router.post(
"/hosts/{host_id}/skills/inventory",
status_code=status.HTTP_204_NO_CONTENT,
)
def report_inventory(
host_id: str,
payload: HostSkillInventoryRequest,
request: Request,
) -> None:
authorize_host(request, host_id)
service.record_host_inventory(
host_id,
json.dumps(payload.skills),
now=utc_now(),
)
return router
# ----------------------------------------------------------------------
# Helpers
# ----------------------------------------------------------------------
def _skill_response(skill) -> CloudSkillResponse:
return CloudSkillResponse(
id=skill.id,
name=skill.name,
kind=skill.kind,
description=skill.description,
tags=list(skill.tags),
revision=skill.revision,
created_at=skill.created_at,
updated_at=skill.updated_at,
content=skill.content,
steps=json.loads(skill.steps_json or "[]"),
parameters=json.loads(skill.parameters_json or "{}"),
)
def _audit(repository, principal: Principal, skill_id: str, action: str) -> None:
repository.record_auth_audit(
AuthAuditEvent(
id=uuid4().hex,
occurred_at=utc_now(),
actor_principal_id=principal.id,
target_user_id=None,
action=action,
outcome="success",
correlation_id=current_correlation_id(),
metadata={"cloud_skill_id": skill_id},
)
)