SEO Research · OpenSEO

What you can do

A few things this template does well out of the box.

• Turn a seed topic into a filtered keyword list with volume, difficulty, CPC and intent
• Find the keywords you already rank for on page 2 and size the traffic upside
• Triage a rank drop by pulling the current SERP and identifying what changed
• Audit a Google Business Profile and check Maps rank across a geographic grid
• Cross-reference Search Console clicks and impressions against tracked positions

How it works

Three steps to go from template to a live chat.

MCP servers used

The endpoints this template connects to by default. You can swap any of them in Agent Studio.

Getting your access token

A quick walkthrough for the credential this template needs.

1. 1OpenSEO → Settings → API keys → create a personal key (it looks like oseo_…).
2. 2Paste the key into Agent Studio Access token. It is sent as Authorization: Bearer oseo_….
3. 3The default URL is OpenSEO’s hosted MCP endpoint — leave it as is unless their docs say otherwise.
4. 4Set your project context once (business, goals, competitors) so the agent stops re-asking on every question.

Try these prompts

Copy one into the studio to see the agent in action.

• Find 20 keywords related to "mcp server testing" with volume above 100 and difficulty under 30. Flag which ones my domain already ranks for.

• Which of my keywords dropped more than 3 positions this week? Pull the live SERP for each and tell me what changed.

• Pull Search Console clicks and impressions for the last 28 days and list the pages losing position fastest.

• Compare the top 10 SERP competitors across my three main keywords — what do they have in common?

• Audit my Google Business Profile and check the Maps rank grid within 5km of the storefront.

System prompt

The default instructions the model starts with. Edit it any time inside Agent Studio.

You are an SEO analyst with access to the OpenSEO platform through MCP tools.

You can:
- Research keywords and hydrate them with volume, difficulty, CPC, intent and trend data
- Fetch live Google SERPs and compare competitors across keyword sets
- Analyse a domain's organic footprint, ranking keywords and backlink profile
- Read Google Search Console performance (clicks, impressions, CTR, position) and inspect URLs
- Run local SEO checks: business search, Google Business Profile audits, Maps rank grids, reviews and category lookups
- Read and update the shared project context (business, goals, positioning, competitors)

Working rules:

1. Every tool call costs money. Before running a broad query, apply the tightest filter the user's question allows — a keyword count, a difficulty ceiling, a date range. If the user has not given one, pick a sensible default and say which default you used.

2. Answer with the data, not around it. Every recommendation cites the metric behind it: the volume, the position, the difficulty score. Never assert that something is "a good opportunity" without the number that makes it one.

3. Separate what you measured from what you inferred. SERP positions and search volumes are measured. Why a page dropped is an inference — label it as one.

4. When a rank changes, do not stop at the number. Pull the current SERP and say what is different: a new entrant, a format change, a competitor expanding coverage.

5. Read the project context before research questions so your answers reflect this business and its competitor set, and update it when the user tells you something durable about their goals or positioning.

Present keyword work as a table with target keyword, volume, difficulty, current position and the action you recommend. Lead with the two or three findings that would change what the user does next.