SEO Data
https://mcp.signaldig.com/data/seo/mcpKeyword overview and intent, related keywords, current SERP observations and Google Trends evidence.
// SignalDig MCP
Connect SignalDig MCP servers to your AI agent
SignalDig MCP gives a compatible AI client authenticated access to focused data and decision tools. It is the capability layer beneath Agent Skills—not a public dashboard and not a replacement for the workflow.
// Endpoints
Choose the server that matches the task
The three product surfaces stay separate so a client can expose only the tools a workflow needs.
Social Data
https://mcp.signaldig.com/data/social/mcpPlatform-native public social retrieval with source context and coverage limitations.
SEO Decisions
https://mcp.signaldig.com/signals/seo/mcpEvidence-constrained keyword and content-opportunity reports with conditions, risks and a next test.
// Client configuration
Send the API key as a Bearer credential
Add only the servers you need to your compatible client's MCP configuration. Field names can vary by client; follow that client's documentation for the exact config file location.
mcp.json
{
"mcpServers": {
"signaldig-seo": {
"type": "http",
"url": "https://mcp.signaldig.com/data/seo/mcp",
"headers": {
"Authorization": "Bearer {SIGNALDIG_API_KEY}"
}
},
"signaldig-social": {
"type": "http",
"url": "https://mcp.signaldig.com/data/social/mcp",
"headers": {
"Authorization": "Bearer {SIGNALDIG_API_KEY}"
}
},
"signaldig-seo-decisions": {
"type": "http",
"url": "https://mcp.signaldig.com/signals/seo/mcp",
"headers": {
"Authorization": "Bearer {SIGNALDIG_API_KEY}"
}
}
}
}Replace locally. Never put a real key in a prompt, repository, screenshot or shared log. SignalDig MCP does not accept an API key in the URL query string.
// Connect
A four-step setup path
- 01
Create an API key
Sign in to SignalDig, create a named API key and save the one-time plaintext value securely.
- 02
Configure one or more servers
Use the production HTTPS endpoint and the Authorization: Bearer header shown above.
- 03
Reload the client
Restart or reload the AI client, then confirm it can list the expected tools before running research.
- 04
Request the smallest useful scope
Use the matching Skill or name a focused task so the agent does not collect unrelated data.
// Current limits
What the connection does—and does not—guarantee
Authentication is required
Missing, expired or revoked API keys are rejected. Keep keys out of browser-readable storage and revoke a key if it may be exposed.
Some work is asynchronous
SEO research and decision reports can return a request ID for polling. Clients should reuse that request and a stable idempotency key instead of submitting duplicates.
Coverage has boundaries
Results depend on the requested market, language, time and available provider coverage. Partial results and missing evidence must remain visible.
No outcome is guaranteed
Market observations and decision support do not guarantee rankings, revenue, reach or product-market fit.
// Why pair MCP with Skills
MCP provides capabilities. Skills show the agent how to use them correctly.
An agent can call MCP tools directly, but a matching Skill reduces guesswork in tool selection and result interpretation. Skills act as an operating handbook for the agent, so the user does not need to learn interface details first.
Select the right tools
Each Skill maps to a defined task and guides the agent to the matching MCP server and tools instead of exploring combinations at random.
Avoid unnecessary calls
A defined research order, scope rules and request reuse keep collection limited to what the current task needs.
Preserve field meaning
Skills require results to retain metric semantics, sources, timing and coverage limits instead of presenting raw numbers without context.
Constrain decision output
The workflows separate retrieval from opportunity decisions and preserve evidence, counter-evidence, risks and a next validation step.