leviathan-public-content 1.0.1__tar.gz

leviathan_public_content-1.0.1/PKG-INFO

@@ -0,0 +1,100 @@
+Metadata-Version: 2.4
+Name: leviathan-public-content
+Version: 1.0.1
+Summary: Official read-only Python client for Leviathan Systems published resources
+Author: Leviathan Systems
+License-Expression: LicenseRef-Proprietary
+Project-URL: Homepage, https://www.leviathansystems.co/developers/sdk
+Project-URL: Documentation, https://www.leviathansystems.co/developers
+Project-URL: Source, https://github.com/leviathansystems-gpu/public-content
+Project-URL: Issues, https://github.com/leviathansystems-gpu/public-content/issues
+Keywords: leviathan-systems,gpu-deployment,public-content
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+
+# Leviathan Systems public content clients 1.0.1
+
+Official dependency-free JavaScript/TypeScript and Python clients for the public,
+read-only content API. Search published titles, follow cursor pages, read a
+catalog resource and retrieve the company profile. No account or API key is
+needed. Package publisher authentication is separate from using the API.
+
+## Source installation
+
+Install from the official source repository:
+
+```sh
+npm install git+https://github.com/leviathansystems-gpu/public-content.git
+python -m pip install git+https://github.com/leviathansystems-gpu/public-content.git
+```
+
+Registry availability is shown in the official SDK documentation. The existing
+website-hosted 1.0.0 archives remain available with their original signatures.
+Version 1.0.1 adds registry metadata and TypeScript declarations; it preserves
+the read-only client behavior. The CLI and Python User-Agent report 1.0.1.
+
+## JavaScript and TypeScript
+
+Use Node.js 20 or later. The JavaScript client also supports modern browsers
+on the same origin. The package includes declarations for TypeScript clients.
+
+```js
+import { LeviathanClient } from 'leviathan-public-content';
+const client = new LeviathanClient();
+const page = await client.listResources({ kind: 'service', q: 'rack', limit: 5 });
+if (page.items.length) {
+  const resource = await client.readResource(page.items[0].id);
+  console.log(resource.canonical, resource.markdown);
+}
+```
+
+The CLI prints JSON. After installation:
+
+```sh
+leviathan-content --help
+leviathan-content search rack integration
+leviathan-content read service:rack-and-stack
+leviathan-content profile
+```
+
+## Python
+
+Use Python 3.10 or later. The client uses only the standard library.
+
+```python
+from leviathan_public_content import LeviathanClient
+client = LeviathanClient()
+page = client.list_resources(kind="service", q="rack", limit=5)
+if page["items"]:
+    resource = client.read_resource(page["items"][0]["id"])
+    print(resource["canonical"], resource["markdown"])
+```
+
+## Limits and errors
+
+Search matches all query words in titles. JavaScript `resourcePages(options)`
+and Python `resource_pages(**options)` retain the original query and origin.
+Restart the search if a content revision invalidates a cursor. A null Markdown
+body means read the returned canonical HTML URL.
+
+`LeviathanAPIError` exposes HTTP status, a stable code and the retry interval
+(`retryAfter` in JavaScript, `retry_after` in Python). Clients do not retry
+automatically. Honor 429 and Retry-After, correct invalid input and keep reads
+bounded. The default request timeout is 15 seconds; redirect responses are not
+followed. Custom origins must be HTTPS, except localhost/127.0.0.1 for local tests.
+
+Educational resources are not a statement of company capabilities, technical
+certification or availability. Read the company profile and confirm the actual
+project scope with the team. These clients cannot read private drafts, send
+messages, book appointments, purchase services or approve engineering work.
+
+## Official references
+
+- https://www.leviathansystems.co/developers
+- https://www.leviathansystems.co/developers/sdk
+- https://www.leviathansystems.co/developers/cli
+- https://www.leviathansystems.co/developers/mcp
+- https://www.leviathansystems.co/openapi.json
+
+The source license remains proprietary; this release does not relicense the
+existing clients. No install scripts or runtime third-party dependencies are used.

leviathan_public_content-1.0.1/README.md

@@ -0,0 +1,59 @@
+# Leviathan Systems agent integration kit
+
+Search published GPU deployment, rack integration and structured cabling
+resources, retrieve canonical documents and read the company's stated service
+scope. No account, API key or directory subscription is required.
+
+## Repository and skill installation
+
+The official integration source is
+https://github.com/leviathansystems-gpu/public-content. Inspect the files before
+installing. To install the resource research skill with the skills CLI:
+
+```sh
+npx skills add leviathansystems-gpu/public-content
+```
+
+This installs guidance; configure the remote MCP endpoint separately if your
+client does not load the portable plugin. A public repository and an install
+command do not establish a skills.sh listing, usage count or platform approval.
+
+## Connect
+
+Add `https://www.leviathansystems.co/mcp` as a remote Streamable HTTP server in
+your MCP client. The endpoint returns JSON and is stateless. Use the client's
+initialization flow, then `tools/list` to inspect the current schemas.
+
+This directory is also an [Agent Plugins 1.0 package](https://agent-plugins.org/specification):
+`plugin.json` describes it, `mcp.json` supplies the remote connection, and
+`skills/leviathan-resource-research/SKILL.md` describes a focused research workflow.
+Extract the versioned ZIP and install the folder through a client that supports
+that format. Field names differ in other clients; use their documented remote
+MCP settings instead of copying the portable manifest into an arbitrary config.
+Installation does not run scripts or create accounts. Connecting sends normal
+public retrieval requests to Leviathan Systems.
+
+## Use the three tools
+
+1. `search_resources({"q":"rack", "kind":"service", "limit":5})`
+2. `read_resource({"id":"service:rack-and-stack"})` using an ID returned by search.
+3. `read_company_profile({"view":"boundaries"})` before assessing service fit.
+
+Search matches every word in the published title. Follow `next_cursor` with the
+original query when needed. A resource without Markdown still has its canonical
+HTML URL. Honor HTTP 429 and `Retry-After`; do not loop over unchanged errors.
+
+Use returned canonical links in citations. Educational articles do not establish
+company capabilities, technical certification or project availability. The tools
+cannot read private drafts, send messages, book appointments, purchase services
+or approve a project. Read [AGENTS.md](./AGENTS.md) and the skill before use.
+
+## Official links
+
+- [Leviathan Systems](https://www.leviathansystems.co/)
+- [MCP documentation](https://www.leviathansystems.co/developers/mcp)
+- [REST API](https://www.leviathansystems.co/developers)
+- [JavaScript and Python clients](https://www.leviathansystems.co/developers/sdk)
+- [CLI](https://www.leviathansystems.co/developers/cli)
+- [MCP.so listing](https://mcp.so/servers/leviathan-systems-public-content-mcp)
+- [Official MCP Registry](https://registry.modelcontextprotocol.io/?q=co.leviathansystems%2Fpublic-content)

leviathan_public_content-1.0.1/SDK.md

@@ -0,0 +1,86 @@
+# Leviathan Systems public content clients 1.0.1
+
+Official dependency-free JavaScript/TypeScript and Python clients for the public,
+read-only content API. Search published titles, follow cursor pages, read a
+catalog resource and retrieve the company profile. No account or API key is
+needed. Package publisher authentication is separate from using the API.
+
+## Source installation
+
+Install from the official source repository:
+
+```sh
+npm install git+https://github.com/leviathansystems-gpu/public-content.git
+python -m pip install git+https://github.com/leviathansystems-gpu/public-content.git
+```
+
+Registry availability is shown in the official SDK documentation. The existing
+website-hosted 1.0.0 archives remain available with their original signatures.
+Version 1.0.1 adds registry metadata and TypeScript declarations; it preserves
+the read-only client behavior. The CLI and Python User-Agent report 1.0.1.
+
+## JavaScript and TypeScript
+
+Use Node.js 20 or later. The JavaScript client also supports modern browsers
+on the same origin. The package includes declarations for TypeScript clients.
+
+```js
+import { LeviathanClient } from 'leviathan-public-content';
+const client = new LeviathanClient();
+const page = await client.listResources({ kind: 'service', q: 'rack', limit: 5 });
+if (page.items.length) {
+  const resource = await client.readResource(page.items[0].id);
+  console.log(resource.canonical, resource.markdown);
+}
+```
+
+The CLI prints JSON. After installation:
+
+```sh
+leviathan-content --help
+leviathan-content search rack integration
+leviathan-content read service:rack-and-stack
+leviathan-content profile
+```
+
+## Python
+
+Use Python 3.10 or later. The client uses only the standard library.
+
+```python
+from leviathan_public_content import LeviathanClient
+client = LeviathanClient()
+page = client.list_resources(kind="service", q="rack", limit=5)
+if page["items"]:
+    resource = client.read_resource(page["items"][0]["id"])
+    print(resource["canonical"], resource["markdown"])
+```
+
+## Limits and errors
+
+Search matches all query words in titles. JavaScript `resourcePages(options)`
+and Python `resource_pages(**options)` retain the original query and origin.
+Restart the search if a content revision invalidates a cursor. A null Markdown
+body means read the returned canonical HTML URL.
+
+`LeviathanAPIError` exposes HTTP status, a stable code and the retry interval
+(`retryAfter` in JavaScript, `retry_after` in Python). Clients do not retry
+automatically. Honor 429 and Retry-After, correct invalid input and keep reads
+bounded. The default request timeout is 15 seconds; redirect responses are not
+followed. Custom origins must be HTTPS, except localhost/127.0.0.1 for local tests.
+
+Educational resources are not a statement of company capabilities, technical
+certification or availability. Read the company profile and confirm the actual
+project scope with the team. These clients cannot read private drafts, send
+messages, book appointments, purchase services or approve engineering work.
+
+## Official references
+
+- https://www.leviathansystems.co/developers
+- https://www.leviathansystems.co/developers/sdk
+- https://www.leviathansystems.co/developers/cli
+- https://www.leviathansystems.co/developers/mcp
+- https://www.leviathansystems.co/openapi.json
+
+The source license remains proprietary; this release does not relicense the
+existing clients. No install scripts or runtime third-party dependencies are used.

leviathan_public_content-1.0.1/leviathan_public_content.egg-info/PKG-INFO

@@ -0,0 +1,100 @@
+Metadata-Version: 2.4
+Name: leviathan-public-content
+Version: 1.0.1
+Summary: Official read-only Python client for Leviathan Systems published resources
+Author: Leviathan Systems
+License-Expression: LicenseRef-Proprietary
+Project-URL: Homepage, https://www.leviathansystems.co/developers/sdk
+Project-URL: Documentation, https://www.leviathansystems.co/developers
+Project-URL: Source, https://github.com/leviathansystems-gpu/public-content
+Project-URL: Issues, https://github.com/leviathansystems-gpu/public-content/issues
+Keywords: leviathan-systems,gpu-deployment,public-content
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+
+# Leviathan Systems public content clients 1.0.1
+
+Official dependency-free JavaScript/TypeScript and Python clients for the public,
+read-only content API. Search published titles, follow cursor pages, read a
+catalog resource and retrieve the company profile. No account or API key is
+needed. Package publisher authentication is separate from using the API.
+
+## Source installation
+
+Install from the official source repository:
+
+```sh
+npm install git+https://github.com/leviathansystems-gpu/public-content.git
+python -m pip install git+https://github.com/leviathansystems-gpu/public-content.git
+```
+
+Registry availability is shown in the official SDK documentation. The existing
+website-hosted 1.0.0 archives remain available with their original signatures.
+Version 1.0.1 adds registry metadata and TypeScript declarations; it preserves
+the read-only client behavior. The CLI and Python User-Agent report 1.0.1.
+
+## JavaScript and TypeScript
+
+Use Node.js 20 or later. The JavaScript client also supports modern browsers
+on the same origin. The package includes declarations for TypeScript clients.
+
+```js
+import { LeviathanClient } from 'leviathan-public-content';
+const client = new LeviathanClient();
+const page = await client.listResources({ kind: 'service', q: 'rack', limit: 5 });
+if (page.items.length) {
+  const resource = await client.readResource(page.items[0].id);
+  console.log(resource.canonical, resource.markdown);
+}
+```
+
+The CLI prints JSON. After installation:
+
+```sh
+leviathan-content --help
+leviathan-content search rack integration
+leviathan-content read service:rack-and-stack
+leviathan-content profile
+```
+
+## Python
+
+Use Python 3.10 or later. The client uses only the standard library.
+
+```python
+from leviathan_public_content import LeviathanClient
+client = LeviathanClient()
+page = client.list_resources(kind="service", q="rack", limit=5)
+if page["items"]:
+    resource = client.read_resource(page["items"][0]["id"])
+    print(resource["canonical"], resource["markdown"])
+```
+
+## Limits and errors
+
+Search matches all query words in titles. JavaScript `resourcePages(options)`
+and Python `resource_pages(**options)` retain the original query and origin.
+Restart the search if a content revision invalidates a cursor. A null Markdown
+body means read the returned canonical HTML URL.
+
+`LeviathanAPIError` exposes HTTP status, a stable code and the retry interval
+(`retryAfter` in JavaScript, `retry_after` in Python). Clients do not retry
+automatically. Honor 429 and Retry-After, correct invalid input and keep reads
+bounded. The default request timeout is 15 seconds; redirect responses are not
+followed. Custom origins must be HTTPS, except localhost/127.0.0.1 for local tests.
+
+Educational resources are not a statement of company capabilities, technical
+certification or availability. Read the company profile and confirm the actual
+project scope with the team. These clients cannot read private drafts, send
+messages, book appointments, purchase services or approve engineering work.
+
+## Official references
+
+- https://www.leviathansystems.co/developers
+- https://www.leviathansystems.co/developers/sdk
+- https://www.leviathansystems.co/developers/cli
+- https://www.leviathansystems.co/developers/mcp
+- https://www.leviathansystems.co/openapi.json
+
+The source license remains proprietary; this release does not relicense the
+existing clients. No install scripts or runtime third-party dependencies are used.

leviathan_public_content-1.0.1/leviathan_public_content.egg-info/SOURCES.txt

@@ -0,0 +1,8 @@
+README.md
+SDK.md
+leviathan_public_content.py
+pyproject.toml
+leviathan_public_content.egg-info/PKG-INFO
+leviathan_public_content.egg-info/SOURCES.txt
+leviathan_public_content.egg-info/dependency_links.txt
+leviathan_public_content.egg-info/top_level.txt

leviathan_public_content-1.0.1/leviathan_public_content.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

leviathan_public_content-1.0.1/leviathan_public_content.egg-info/top_level.txt

@@ -0,0 +1 @@
+leviathan_public_content

leviathan_public_content-1.0.1/leviathan_public_content.py

@@ -0,0 +1,85 @@
+"""Official dependency-free client for Leviathan Systems public content."""
+import json
+import urllib.error
+import urllib.parse
+import urllib.request
+
+class LeviathanAPIError(Exception):
+    def __init__(self, message, status=0, code="NETWORK_ERROR", retry_after=None):
+        super().__init__(message)
+        self.status, self.code, self.retry_after = status, code, retry_after
+
+class _NoRedirect(urllib.request.HTTPRedirectHandler):
+    def redirect_request(self, req, fp, code, msg, headers, newurl):
+        return None
+
+class LeviathanClient:
+    def __init__(self, base_url="https://www.leviathansystems.co", timeout=15):
+        url = urllib.parse.urlsplit(base_url)
+        local = url.scheme == "http" and url.hostname in ("localhost", "127.0.0.1")
+        if url.username or url.password or url.query or url.fragment or url.path not in ("", "/") or not url.hostname or not (url.scheme == "https" or local):
+            raise ValueError("base_url must be an HTTPS origin, or a local development origin")
+        if not isinstance(timeout, (int, float)) or not 0 < timeout <= 60:
+            raise ValueError("timeout must be from 0 to 60 seconds")
+        self.base_url, self.timeout = base_url.rstrip("/"), timeout
+        self._opener = urllib.request.build_opener(_NoRedirect)
+
+    def _request(self, path):
+        req = urllib.request.Request(self.base_url + path, headers={"Accept": "application/json", "Leviathan-API-Version": "1", "User-Agent": "Leviathan-Public-Content-Python/1.0.1"})
+        try:
+            try:
+                response = self._opener.open(req, timeout=self.timeout)
+            except urllib.error.HTTPError as error:
+                response = error
+            with response:
+                raw = response.read(2_000_001)
+                if len(raw) > 2_000_000:
+                    raise LeviathanAPIError("Response exceeds the client size limit", code="INVALID_RESPONSE")
+                try:
+                    data = json.loads(raw)
+                except (ValueError, UnicodeError):
+                    raise LeviathanAPIError("The server returned an unreadable response", response.status, "INVALID_RESPONSE") from None
+                if not isinstance(data, dict):
+                    raise LeviathanAPIError("The server returned an invalid document", response.status, "INVALID_RESPONSE")
+                if not 200 <= response.status < 300:
+                    raise LeviathanAPIError(data.get("detail", "Content request failed"), response.status, data.get("code", "HTTP_ERROR"), response.headers.get("Retry-After"))
+                return data
+        except (urllib.error.URLError, TimeoutError, OSError):
+            raise LeviathanAPIError("The content request could not complete. Check connectivity and retry.") from None
+
+    def list_resources(self, q=None, kind=None, limit=20, cursor=None):
+        if q is not None and (not isinstance(q, str) or len(q) > 200):
+            raise ValueError("q must be a string of at most 200 characters")
+        if kind is not None and kind not in ("article", "service"):
+            raise ValueError("kind must be article or service")
+        if type(limit) is not int or not 1 <= limit <= 50:
+            raise ValueError("limit must be from 1 to 50")
+        if cursor is not None and (not isinstance(cursor, str) or not 1 <= len(cursor) <= 100 or any(c not in "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789_-" for c in cursor)):
+            raise ValueError("cursor must come from a previous resource page")
+        params = {"limit": limit}
+        for key, value in (("q", q), ("kind", kind), ("cursor", cursor)):
+            if value is not None:
+                params[key] = value
+        return self._request("/api/v1/resources?" + urllib.parse.urlencode(params))
+
+    def resource_pages(self, **options):
+        cursor = options.pop("cursor", None)
+        seen = set()
+        while True:
+            page = self.list_resources(cursor=cursor, **options)
+            yield page
+            cursor = page.get("next_cursor")
+            if not cursor:
+                return
+            if cursor in seen:
+                raise LeviathanAPIError("The server repeated a cursor", code="INVALID_RESPONSE")
+            seen.add(cursor)
+
+    def read_resource(self, resource_id):
+        import re
+        if not isinstance(resource_id, str) or len(resource_id) > 210 or not re.fullmatch(r"(?:article|service):[a-z0-9-]+", resource_id):
+            raise ValueError("Use an article:slug or service:slug catalog ID")
+        return self._request("/api/v1/resources/" + urllib.parse.quote(resource_id, safe=""))
+
+    def company_profile(self):
+        return self._request("/api/machine/home")

leviathan_public_content-1.0.1/pyproject.toml

@@ -0,0 +1,23 @@
+[build-system]
+requires = ["setuptools>=68"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "leviathan-public-content"
+version = "1.0.1"
+description = "Official read-only Python client for Leviathan Systems published resources"
+readme = "SDK.md"
+requires-python = ">=3.10"
+license = "LicenseRef-Proprietary"
+authors = [{name = "Leviathan Systems"}]
+keywords = ["leviathan-systems", "gpu-deployment", "public-content"]
+dependencies = []
+
+[project.urls]
+Homepage = "https://www.leviathansystems.co/developers/sdk"
+Documentation = "https://www.leviathansystems.co/developers"
+Source = "https://github.com/leviathansystems-gpu/public-content"
+Issues = "https://github.com/leviathansystems-gpu/public-content/issues"
+
+[tool.setuptools]
+py-modules = ["leviathan_public_content"]

leviathan_public_content-1.0.1/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+