arsenkin-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| PUBMED_EMAIL | Yes | Your email address (required by NCBI) | |
| PUBMED_API_KEY | No | Optional API key for higher rate limits |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| logging | {} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| extensions | {
"io.modelcontextprotocol/ui": {}
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| arsenkin_profile_catalogA | List supported profiles, payload fields and example SEO requests; no provider calls. |
| arsenkin_local_diagnosticsA | Explain local worker heartbeat, your queue and configured budget; no provider calls. |
| arsenkin_result_tableA | Create SERP URL/host frequency or cluster tables from a selected saved-result JSON pointer. serp_urls/serp_hosts select repeatUrl/repeatHost; clusters selects cluster records. Completeness covers only this selected view, never the whole provider task. No new paid calls. |
| arsenkin_statusA | Check Arsenkin account/API status; this does not submit a paid task. |
| arsenkin_estimateB | Validate an Arsenkin payload and estimate cost without a live call. |
| arsenkin_batch_estimateB | Plan, coalesce and price a multi-request batch without a live call. |
| arsenkin_request_hash_lookupA | Look up exact server-ledger hashes without exposing job or provider payload data. |
| arsenkin_prepareC | Create a bounded approval; trusted policy may return typed server acceptance. |
| arsenkin_batch_prepareC | Create one bounded batch approval; trusted policy may accept it server-side. |
| arsenkin_submitC | Queue an approved paid Arsenkin request; never blocks on provider polling. |
| arsenkin_batch_submitC | Queue one approved batch job; worker coalesces and polls centrally. |
| arsenkin_task_statusB | Return durable status for the principal's Arsenkin job. |
| arsenkin_get_resultC | Return the durable result for the principal's Arsenkin job. |
| arsenkin_export_resultA | Export a saved owned job as bounded JSONL/CSV with JSON-pointer provenance; no provider calls. |
| arsenkin_cancelB | Request cancellation of the principal's queued or running Arsenkin job. |
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 15 tools
Most tools occupy distinct workflow stages (estimate, prepare, submit, status, result, cancel), and single/batch variants are clearly labeled. A few near pairs like arsenkin_status vs arsenkin_task_status and get_result vs export_result could confuse an agent, though the descriptions do separate them.
All names share the arsenkin_ prefix and snake_case, and paired batch tools follow a clear modifier pattern. However, conventions are mixed: nouns like status and profile_catalog sit alongside bare verbs like prepare, submit, and cancel, plus get_/export_ prefixes in get_result and export_result, so the set is readable but not uniformly verb_noun.
Fifteen tools is on the upper end of a well-scoped set, but every tool addresses a distinct stage of the paid-task lifecycle, including estimates, approvals, submission, retrieval, export, and cancellation. Batch variants add count but are justified by the server's explicit batch workflow.
The core lifecycle is well covered: estimate, prepare, submit, task_status, get_result, export_result, and cancel, for both single and batch paths. The main gaps are lack of a general job-list or discovery tool and no batch-specific status/cancel endpoint, but agents can likely work around these.