diff --git a/scrapinghub/__init__.py b/scrapinghub/__init__.py
index 39a7fe97..530c8c50 100644
--- a/scrapinghub/__init__.py
+++ b/scrapinghub/__init__.py
@@ -1,4 +1,4 @@
-__all__ = ["APIError", "Connection", "HubstorageClient"]
+__all__ = ["APIError", "Connection", "HubstorageClient", "ScrapinghubClient"]
 
 
 import pkgutil
@@ -9,3 +9,4 @@
 
 from .legacy import *
 from .hubstorage import HubstorageClient
+from .client import ScrapinghubClient
diff --git a/scrapinghub/client.py b/scrapinghub/client.py
new file mode 100644
index 00000000..19465219
--- /dev/null
+++ b/scrapinghub/client.py
@@ -0,0 +1,352 @@
+import logging
+
+from requests.compat import urlencode
+from requests.compat import urljoin
+
+from scrapinghub.hubstorage import HubstorageClient
+from scrapinghub.hubstorage.jobq import DuplicateJobError
+from scrapinghub.hubstorage.project import Ids
+from scrapinghub.hubstorage.project import Reports
+from scrapinghub.hubstorage.project import Samples as ProjectSamples
+from scrapinghub.hubstorage.project import Settings
+from scrapinghub.hubstorage.activity import Activity
+from scrapinghub.hubstorage.collectionsrt import Collections
+from scrapinghub.hubstorage.frontier import Frontier
+from scrapinghub.hubstorage.utils import urlpathjoin, xauth
+
+# import the classes to redefine them in the module
+from scrapinghub.hubstorage import resourcetype
+from scrapinghub.hubstorage.client import Projects as _Projects
+from scrapinghub.hubstorage.project import Project as _Project
+from scrapinghub.hubstorage.project import Spiders as _Spiders
+from scrapinghub.hubstorage.job import Job as _Job
+from scrapinghub.hubstorage.job import Items as _Items
+from scrapinghub.hubstorage.job import Logs as _Logs
+from scrapinghub.hubstorage.job import Requests as _Requests
+from scrapinghub.hubstorage.job import JobMeta as _JobMeta
+from scrapinghub.hubstorage.job import Samples as _Samples
+from scrapinghub.hubstorage.jobq import JobQ as _JobQ
+
+from scrapinghub.hubstorage.serialization import jldecode, mpdecode
+
+
+class ScrapinghubAPIError(Exception):
+    pass
+
+
+class Log(object):
+    DEBUG = logging.DEBUG
+    INFO = logging.INFO
+    WARNING = logging.WARNING
+    ERROR = logging.ERROR
+    CRITICAL = logging.CRITICAL
+    SILENT = CRITICAL + 1
+
+
+class ScrapinghubClient(HubstorageClient):
+
+    DEFAULT_DASH_ENDPOINT = 'https://app.scrapinghub.com/api/'
+
+    def __init__(self, auth=None, endpoint=None, dash_endpoint=None,
+                 connection_timeout=None, max_retries=None,
+                 max_retry_time=None, user_agent=None):
+        self.auth = xauth(auth)
+        self.endpoint = endpoint or self.DEFAULT_ENDPOINT
+        self.connection_timeout = (connection_timeout or
+                                   self.DEFAULT_CONNECTION_TIMEOUT_S)
+        self.user_agent = user_agent or self.DEFAULT_USER_AGENT
+        self.session = self._create_session()
+        self.retrier = self._create_retrier(max_retries, max_retry_time)
+        self.jobq = JobQ(self, None)
+        self.projects = Projects(self, None)
+        self.root = ResourceType(self, None)
+        self.dash_endpoint = dash_endpoint or self.DEFAULT_DASH_ENDPOINT
+        self._batchuploader = None
+
+    def get_job(self, *args, **kwargs):
+        return Job(self, *args, **kwargs)
+
+    def push_job(self, *args, **kwargs):
+        raise AttributeError(
+            "Scheduling jobs from client level is deprecated."
+            "Please schedule new jobs via project.push_job()."
+        )
+
+    def project_ids(self):
+        """Returns a list of available projects."""
+        return self.projects.list()
+
+# ------------------ special dash mixin class ------------------
+
+
+class DashMixin(object):
+    """A simple mixin class to work with Dash API endpoints"""
+
+    def _dash_apiget(self, endpoint, basepath, params=None, raw=False,
+                     auth=None, **kwargs):
+        """Performs GET request to Dash endpoint.
+        Thin wrapper over requests.session.get.
+        """
+        url = urljoin(endpoint, basepath)
+        if params:
+            url = "{0}?{1}".format(url, urlencode(params, True))
+        response = self.client.session.get(
+            url, auth=auth or self.auth, **kwargs)
+        return self._decode_dash_response(response, raw)
+
+    def _dash_apipost(self, endpoint, basepath, raw=False, auth=None,
+                      **kwargs):
+        """Performs POST request to Dash endpoint.
+        Thin wrapper over requests.session.post.
+        """
+        url = urljoin(endpoint, basepath)
+        response = self.client.session.post(
+            url, auth=auth or self.auth, **kwargs)
+        return self._decode_dash_response(response, raw)
+
+    def _decode_dash_response(self, response, raw):
+        if raw:
+            return response.raw
+        response.raise_for_status()
+        data = response.json()
+        try:
+            if data['status'] == 'ok':
+                return data
+            elif data['status'] in ('error', 'badrequest'):
+                raise ScrapinghubAPIError(data['message'])
+            raise ScrapinghubAPIError(
+                "Unknown response status: {0[status]}".format(data))
+        except KeyError:
+            raise ScrapinghubAPIError("JSON response does not contain status")
+
+
+# ------------------ resource type classes section ------------------
+
+class ResourceType(resourcetype.ResourceType, DashMixin):
+
+    def __init__(self, client, key, auth=None):
+        super(ResourceType, self).__init__(client, key, auth=auth)
+        self._key = key
+
+
+class MappingResourceType(resourcetype.MappingResourceType, ResourceType):
+    """Custom MappingResourceType based on modified ResourceType.
+    MRO: 1) MappingResourceType
+         2) hubstorage.resourcetype.MappingResourceType
+         3) ResourceType
+         4) hubstorage.resourcetype.ResourceType"""
+
+
+class ItemsResourceType(resourcetype.ItemsResourceType, ResourceType):
+    """Custom ItemsResourceType based on modified ResourceType
+    MRO: 1) ItemsResourceType
+         2) hubstorage.resourcetype.ItemsResourceType
+         3) ResourceType
+         4) hubstorage.resourcetype.ResourceType"""
+
+    def list(self, _key=None, **params):
+        if hasattr(self, 'iter'):
+            return list(self.iter(_key=_key, **params))
+        return super(ItemsResourceType, self).list(_key=None, **params)
+
+
+class DownloadableResource(resourcetype.DownloadableResource):
+    """Custom DownloadableResource based on modified ResourceType
+    MRO: 1) DownloadableResource
+         2) hubstorage.resourcetype.DownloadableResource
+         3) ResourceType
+         4) hubstorage.resourcetype.ResourceType
+    """
+    def __init__(self, *args, **kwargs):
+        super(DownloadableResource, self).__init__(*args, **kwargs)
+        self.iter_raw_msgpack = super(DownloadableResource, self).iter_msgpack
+        self.iter_raw_json = super(DownloadableResource, self).iter_json
+
+    def iter(self, *args, **kwargs):
+        """Reliably iterate through all data as python objects
+
+        calls either iter_json or iter_msgpack, decoding the results
+        """
+        if self._allows_mpack():
+            return mpdecode(self.iter_raw_msgpack(*args, **kwargs))
+        return jldecode(self.iter_raw_json(*args, **kwargs))
+
+    def __getattribute__(self, name):
+        rename_map = {
+            'iter_values': 'iter',
+            'iter_json': 'iter_raw_json',
+            'iter_msgpack': 'iter_raw_msgpack',
+        }
+        if name in rename_map:
+            raise AttributeError(
+                'Method was renamed to %s()' % rename_map[name])
+        return super(DownloadableResource, self).__getattribute__(name)
+
+
+# ------------------ project classes section ---------------------
+
+
+class Projects(_Projects, ResourceType):
+
+    def get(self, *args, **kwargs):
+        return Project(self.client, *args, **kwargs)
+
+    def list(self):
+        return self._dash_apiget(self.client.dash_endpoint,
+                                 'scrapyd/listprojects.json')
+
+
+class Project(_Project):
+
+    def __init__(self, client, projectid, auth=None):
+        self.client = client
+        self.projectid = urlpathjoin(projectid)
+        assert len(self.projectid.split('/')) == 1, \
+            'projectkey must be just one id: %s' % projectid
+        self.auth = xauth(auth) or client.auth
+        self.items = Items(client, self.projectid, auth=auth)
+        self.logs = Logs(client, self.projectid, auth=auth)
+        self.samples = ProjectSamples(client, self.projectid, auth=auth)
+        self.jobq = JobQ(client, self.projectid, auth=auth)
+        self.activity = Activity(client, self.projectid, auth=auth)
+        self.collections = Collections(client, self.projectid, auth=auth)
+        self.frontier = Frontier(client, self.projectid, auth=auth)
+        self.ids = Ids(client, self.projectid, auth=auth)
+        self.settings = Settings(client, self.projectid, auth=auth)
+        self.reports = Reports(client, self.projectid, auth=auth)
+        self.spiders = Spiders(client, self.projectid, auth=auth)
+
+    def push_job(self, spidername, **jobparams):
+        data = self.jobq.push(spidername, **jobparams)
+        key = data['key']
+        return Job(self.client, key, auth=self.auth)
+
+    def count(self, **params):
+        return self.jobq.count(**params)
+
+
+class Spiders(_Spiders, ResourceType):
+
+    def list(self):
+        return self._dash_apiget(self.client.dash_endpoint,
+                                 'spiders/list.json',
+                                 params={'project': self._key})['spiders']
+
+    def update_tags(self, spidername, add=None, remove=None):
+        params = get_tags_for_update(add_tag=add, remove_tag=remove)
+        if not params:
+            return
+        params.update({'spider': "noop", 'project': self._key})
+        return self._dash_apipost(self.client.dash_endpoint,
+                                  'jobs/update.json', data=params)['count']
+
+
+# ------------------------ job-related section -----------------------
+
+class Job(_Job):
+
+    def __init__(self, client, key, auth=None, jobauth=None, metadata=None):
+        self.key = urlpathjoin(key)
+        assert len(self.key.split('/')) == 3, \
+            'Jobkey must be projectid/spiderid/jobid: %s' % self.key
+        self.jobauth = jobauth
+        self.auth = self.jobauth or auth
+        self.metadata = JobMeta(client, self.key, self.auth, cached=metadata)
+        self.items = Items(client, self.key, self.auth)
+        self.logs = Logs(client, self.key, self.auth)
+        self.samples = Samples(client, self.key, self.auth)
+        self.requests = Requests(client, self.key, self.auth)
+        self.jobq = JobQ(client, self.key.split('/')[0], auth)
+
+    def update_tags(self, *args, **kwargs):
+        return self.metadata.update_tags(*args, **kwargs)
+
+
+class JobMeta(_JobMeta, MappingResourceType):
+
+    def update_tags(self, add=None, remove=None):
+        params = get_tags_for_update(add_tag=add, remove_tag=remove)
+        if not params:
+            return
+        params['job'] = self._key
+        params['project'] = self._key.split('/', 1)[0]
+        response = self._dash_apipost(self.client.dash_endpoint,
+                                      'jobs/update.json', data=params)
+        return response['count']
+
+
+class Items(_Items, ItemsResourceType, DownloadableResource):
+
+    def iter(self, _key=None, **params):
+        if 'offset' in params:
+            params['start'] = '%s/%s' % (self._key, params['offset'])
+            del params['offset']
+        return self.apiget(_key, params=params)
+
+
+class Logs(_Logs, ItemsResourceType, DownloadableResource):
+
+    def iter(self, _key=None, **params):
+        if 'offset' in params:
+            params['start'] = '%s/%s' % (self._key, params['offset'])
+            del params['offset']
+        if 'level' in params:
+            minlevel = getattr(Log, params.get('level'), None)
+            if minlevel is None:
+                raise ScrapinghubAPIError(
+                    "Unknown log level: %s" % params.get('level'))
+            params['filters'] = ['level', '>=', [minlevel]]
+        return self.apiget(_key, params=params)
+
+
+class Requests(_Requests, ItemsResourceType, DownloadableResource):
+
+    def iter(self, _key=None, **params):
+        return self.apiget(_key, params=params)
+
+
+class Samples(_Samples, ItemsResourceType):
+
+    def iter(self, _key=None, **params):
+        return self.apiget(_key, params=params)
+
+# ------------------------ jobq section -----------------------
+
+
+class JobQ(_JobQ, ResourceType):
+
+    def count(self, **params):
+        return next(self.apiget(('count',), params=params))
+
+    def push(self, spider, **jobparams):
+        jobparams['spider'] = spider
+        # for client-level JobQ, project should be provided via jobparams
+        if 'project' not in jobparams:
+            if not self._key:
+                raise ScrapinghubAPIError(
+                    "Project is required when scheduling new jobs")
+            jobparams['project'] = self._key
+        # FIXME JobQ endpoint can schedule multiple jobs with json-lines,
+        # corresponding Dash endpoint - only one job per request
+        response = self._dash_apipost(self.client.dash_endpoint,
+                                      'schedule.json',
+                                      data=jobparams)
+        if response.get('status') == 'error':
+            if 'already scheduled' in response['message']:
+                raise DuplicateJobError(response['message'])
+            raise ScrapinghubAPIError(response['message'])
+        return {'key': response.get('jobid'), 'auth': None}
+
+
+# ------------------------ auxiliaries section -----------------------
+
+def get_tags_for_update(**kwargs):
+    """Helper to check tags changes"""
+    params = {}
+    for k, v in kwargs.items():
+        if not v:
+            continue
+        if not isinstance(v, list):
+            raise ScrapinghubAPIError("Add/remove value must be a list")
+        params[k] = v
+    return params