Xzar MCP Server
OfficialServer Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| XZAR_API_KEY | Yes | Your Xzar API key. Get one with $5 of free credit at https://xzar.si/signup.html. |
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 | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| start_taskB | Start a new task episode. Returns the episode id, the instructions and the tool budget. |
| search_libraryC | Search the episode's prior-art library (publications dated before the filing date). |
| read_publicationC | Read one publication from the episode's library. |
| search_rulesC | Search the examination manual for the rules and tests examiners apply. |
| read_ruleC | Read one examination manual section by number, e.g. 2143.01. |
| submit_answerA | Submit the final answer (JSON object in the format the instructions give). One per episode. |
| get_scorecardB | Your scorecard: scores by skill and where your agent fails. |
| get_usageB | Your credit balance and recent charges. |
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 8 tools
Each tool has a clearly distinct purpose: task lifecycle (start_task, submit_answer), library access (search_library, read_publication), rules access (search_rules, read_rule), and feedback/usage (get_scorecard, get_usage). No two tools overlap in function.
All tool names follow a consistent snake_case verb_noun pattern (start_task, search_library, read_publication, search_rules, read_rule, submit_answer, get_scorecard, get_usage). The convention is applied uniformly with no deviations.
Eight tools is well-scoped for an agent task environment. It covers the essential operations without redundancy, and each tool earns its place by serving a distinct step in the workflow.
The tool surface covers the full episode lifecycle: starting a task, researching prior art and rules, submitting an answer, and reviewing feedback and usage. No critical operations appear to be missing for the stated purpose.