GEO MCP by DigestSEO
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
Use this server to measure stored AI visibility, not to infer fresh results. If the brand_id is unknown, call list_brands first. For a new local brand, call track_brand, then refresh_brand, then check_visibility. If a tracked brand's domain, name, category, competitors, aliases, exclusions, or refresh cadence changes, use update_brand so prompts and historical runs stay intact. Local refresh_brand is synchronous. Use citation, competitor, history, and content-gap tools only after scan data exists. Engines without configured API keys are unavailable and must not be treated as zero visibility.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| check_visibilityA | Read stored visibility scores. run_mode='live' or 'batch' selects stored scans; omit for latest across modes. Live does not guarantee citations. Example: brand_id='acme', engines=['chatgpt']. |
| get_visibility_historyA | Daily/weekly visibility history, with model-only versus search-requested evidence. current_identity_only=true filters known old-brand runs. Example: brand_id='acme', days=30, granularity='weekly'. |
| compare_competitorsB | Compare brand/competitor mentions and share of voice. active_prompts_only=true excludes retired prompts; default keeps history. Unmeasured coverage and tracking caveats are reported. Example: brand_id='acme', days=7. |
| get_citationsA | Return paginated brand mentions, link-only evidence, request mode per row, and top source domains. Search requests do not guarantee citations. Reuse next_cursor within 24 hours. Example: brand_id='acme', days=14. |
| get_content_gapsB | Prioritize observed competitor-winning prompts. Provider-free by default; ai_analysis=true opts in to token-billed Claude (may retry once). Example: brand_id='acme'. |
| refresh_brandA | Scan selected engines and return after completion. Provider API requests may incur charges. Retries rerun every active prompt, including prior successes. Example: brand_id='acme', engines=['chatgpt','gemini']. |
| track_brandA | Track a brand with 3 provider-free starter prompts. Set prompt_generation='anthropic' for paid Claude generation (may retry). Example: brand_id='acme', name='Acme', domain='acme.com'. |
| update_brandA | Change tracked-brand metadata without replacing prompts or history. Example: brand_id='acme', competitors=['rival.com'], refresh_frequency='manual' pauses cron scans but keeps manual refresh. |
| list_brandsA | List tracked brands with metadata and active-prompt counts. Example: call before other tools when you need a brand_id. |
| list_promptsA | List a brand's active buyer-intent prompts without changing them. Example: brand_id='acme' before an audit or refresh. |
| set_promptsA | Replace a brand's active prompts with an exact supplied set; history is preserved. Example: brand_id='acme', prompts=['What is the best X for Y?']. |
| generate_promptsA | Replace active prompts with Claude-Haiku-generated buyer-intent prompts; requires ANTHROPIC_API_KEY and preserves history. Example: brand_id='acme', count=20. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 12 tools
Most tools have clearly distinct purposes: citations, visibility scores, history, competitor comparison, content gaps, brand management, and prompt management are separable. The main potential confusion is between check_visibility and get_visibility_history, and between refresh_brand and check_visibility, but descriptions clarify stored vs live and current vs historical.
All tool names use snake_case with a consistent verb_noun pattern (get_, check_, compare_, refresh_, track_, update_, list_, generate_, set_). The verbs vary appropriately by action, and no naming convention is mixed.
Twelve tools is well within the ideal 3-15 range for a brand visibility tracking server. Each tool covers a distinct operation (read, compare, refresh, manage brands/prompts) and none appear redundant.
The surface covers core workflows: brand creation/update/listing, prompt listing/generation/replacement, visibility checks/history, citations, competitor comparison, content gaps, and refresh. Minor gaps exist, such as no explicit delete_brand or get_brand tool, but agents can work around these via list_brands and update_brand.