Server data from the Official MCP Registry
Splunk Observability Cloud MCP server: traces, metrics, alerts, and dashboards for IDEs.
Splunk Observability Cloud MCP server: traces, metrics, alerts, and dashboards for IDEs.
Remote endpoints: streamable-http: https://api.{realm}.signalfx.com/v2/mcp
Valid MCP server (1 strong, 1 medium validity signals). No known CVEs in dependencies. Imported from the Official MCP Registry. 1 finding(s) downgraded by scanner intelligence.
Endpoint verified · Open access · 1 issue found
Security scores are indicators to help you make informed decisions, not guarantees. Always review permissions before connecting any MCP server.
This plugin requests these system permissions. Most are normal for its category.
Remote Plugin
No local installation needed. Your AI client connects to the remote endpoint directly.
Add this to your MCP configuration to connect:
{
"mcpServers": {
"io-github-signalfx-splunk-o11y-mcp-connect": {
"url": "https://api.{realm}.signalfx.com/v2/mcp"
}
}
}From the project's GitHub README.
Connects AI coding IDEs (VS Code, Cursor, Codex, Claude Code, Kiro) to the Splunk Observability remote MCP server.
npx @splunk/o11y-mcp-connect connect
You'll be prompted for your realm (e.g. us0, eu0) and an access token. The command writes the
splunk-o11y-mcp server entry into each supported IDE's user-scope config and verifies the connection with a
tools/list call before exiting.
The extension in extensions/vscode registers an McpServerDefinitionProvider so VS Code manages
the MCP server directly — no mcp.json file, and no npx needed:
npm install && npm run build at the repo root first — the extension's file:../..
dependency only exposes what's in dist/, so the root must be built before the extension is.npm install && npm run build inside extensions/vscode.The token is verified against the endpoint and, only once verified, stored in VS Code's
SecretStorage (OS keychain) rather than a plaintext config file. VS Code calls back into the
extension to inject the token header each time it starts the server, so the token never sits in
mcp.json or any other file on disk. Run Splunk Observability: Remove Stored Credentials to
delete it. Marketplace publishing is deferred to follow-up work.
Generate a one-click install link for Cursor without going through the interactive wizard:
npx @splunk/o11y-mcp-connect cursor-link us0
This prints a cursor://anysphere.cursor-deeplink/mcp/install?... link. The link carries only the
endpoint URL — never a token — so after clicking it, add your access token inside Cursor to finish
connecting.
Generate a one-click install link for VS Code without going through the interactive wizard:
npx @splunk/o11y-mcp-connect vscode-link us0
This prints a vscode:mcp/install?... link. The link carries only the endpoint URL — never a
token — so after clicking it, add your access token inside VS Code to finish connecting.
Install the plugin from this repo's self-hosted marketplace instead of running the wizard by hand:
/plugin marketplace add signalfx/splunk-o11y-mcp-connect
/plugin install splunk-o11y-mcp@splunk-o11y-mcp-connect
Set SPLUNK_O11Y_REALM and SPLUNK_O11Y_TOKEN in the shell Claude Code runs in, then run the
connect skill — it never asks for the token in chat, since anything typed into chat lands in the
conversation transcript. This only configures Claude Code (equivalent to connect --ide claude-code) — for the other IDEs, use npx @splunk/o11y-mcp-connect connect directly.
Codex has no marketplace or one-click install link, so the command is its only front door:
npx @splunk/o11y-mcp-connect connect --ide codex
This writes the splunk-o11y-mcp server into ~/.codex/config.toml as a url +
http_headers entry under a managed block, alongside any existing Codex MCP config.
Kiro has no one-click install link or CLI front door yet, so the connect command is its only
front door:
npx @splunk/o11y-mcp-connect connect --ide kiro
This writes the splunk-o11y-mcp server into ~/.kiro/settings/mcp.json under mcpServers,
the same shape as the Cursor adapter.
By default, connect configures all five supported IDEs. Pass --ide to configure a subset —
a single IDE, or a comma-separated list:
npx @splunk/o11y-mcp-connect connect --ide claude-code
npx @splunk/o11y-mcp-connect connect --ide cursor,codex
Set SPLUNK_O11Y_REALM and SPLUNK_O11Y_TOKEN in the environment before running connect to
skip the interactive prompts entirely (both must be set together):
SPLUNK_O11Y_REALM=us0 SPLUNK_O11Y_TOKEN=<token> npx @splunk/o11y-mcp-connect connect
If Node/npx isn't available, see the manual config snippets in docs/manual-setup.md.
[FAILED] ... Token was rejected — the access token was rejected by the server (HTTP
401/403). Mint a new token in Splunk Observability Cloud and re-run connect.[FAILED] ... Server returned <status> — the endpoint responded but not successfully.
Double-check the realm you entered; re-run connect after confirming it.[FAILED] ... Could not reach https://api.<realm>.signalfx.com/v2/mcp — a network-level
failure (DNS, firewall, VPN). Confirm the realm is spelled correctly and that you have
connectivity to *.signalfx.com.claude mcp add failed — the Claude Code CLI isn't installed, isn't on your PATH, or is
out of date. Install/update it and re-run connect.connect in an
interactive terminal (or piping realm/token as two newline-separated lines on stdin, in that
order, for non-interactive use).connect succeeded — the connector runs once and exits; it can't
detect expiry after the fact. You'll see this as a failed tool call from within the IDE
itself. To reconnect: mint a fresh token and re-run the same front door you used originally
(connect, the Cursor link, or the Claude Code plugin) — it overwrites the existing entry.[OK]/[FAILED] output
(redact your token first).Be the first to review this server!
by Modelcontextprotocol · Developer Tools
Web content fetching and conversion for efficient LLM usage
by Toleno · Developer Tools
Toleno Network MCP Server — Manage your Toleno mining account with Claude AI using natural language.
by mcp-marketplace · Developer Tools
Create, build, and publish Python MCP servers to PyPI — conversationally.