A local MCP server for publishing HTML files to StelaSpace from coding agents that work with your local files — Claude Code and Codex.
It runs on your machine, so it reads HTML files directly from disk by path and uploads the bytes via the StelaSpace API. That's what lets it publish large (1 MB+) reports and dashboards — unlike passing file contents as a tool argument, which is capped by the model's output size.
Claude Desktop (chat) isn't supported for publishing: its sandbox can't share files with a local MCP server, so it can't upload local reports. Use Claude Code or Codex, where the agent generates the report to disk and publishes it directly.
publish_file — { path, spaceSlug, title?, slug?, tags?, visibility? } → reads
the file and returns its permanent URL. Reusing a slug publishes a new version.
list_spaces — list the spaces in your team.
list_documents — { spaceSlug, query? } — list documents in a space.
Related MCP server: deloc-mcp
Get an API key from StelaSpace → Settings → API Keys (starts with ss_sk_).
claude mcp add stelaspace --scope user \
--env STELASPACE_API_KEY=ss_sk_your_key \
-- npx -y stelaspace-mcp~/.codex/config.toml:
[mcp_servers.stelaspace]
command = "npx"
args = ["-y", "stelaspace-mcp"]
env = { STELASPACE_API_KEY = "ss_sk_your_key" }Then ask the agent to build and publish a report, e.g. "generate the sales dashboard and publish the HTML to my reports space as public."
STELASPACE_API_KEY (required) — your ss_sk_ key.
STELASPACE_API_URL (optional) — override the API endpoint. Defaults to
https://stelaspace.com; most users never need to set this.
List documents in a space, optionally filtered by a case-insensitive title search term.
Inputs: spaceSlug (required), query.
List the spaces in your team. Use a space's slug as the spaceSlug when publishing a document.
Inputs: no named parameters.
Upload a local .html/.htm file to a StelaSpace space and get back a permanent, shareable URL. The file is read from disk by path, so large files work fine. Reusing a slug publishes a new version.
Inputs: path (required), spaceSlug (required), slug, tags, +2 more.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 3 tools
Each tool has a distinct purpose: listing documents, listing spaces, and publishing files. There is no overlap between their functionalities.
All tool names follow a consistent verb_noun pattern: list_documents, list_spaces, and publish_file. The naming is clear and predictable.
With 3 tools, the server is well-scoped for a simple document publishing service. The count falls within the ideal 3-15 range, and each tool serves a clear role.
The tool set covers listing and publishing but lacks essential operations like deleting documents or retrieving a single document's details. These gaps could hinder common workflows.
Shared memory MCP server for Claude Code, Codex, Gemini CLI and Cursor. Cited, editable.
Persistent file storage for AI agents. Save files over MCP, get a link a human can open.
MarkupBase turns AI-generated Markdown and HTML into durable, versioned artifacts that people can review and discuss. Its MCP server lets agents publish new versions, preserve contextual comments, include hosted images, and respond to feedback through secure account-linked identities, creating a clear human review boundary without requiring real-time editing.
Deploy static sites and connect custom domains from Claude, Cursor, or any MCP agent.
