MCP Agent Toolkit
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.
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 |
|---|---|
| blackboard_writeB | Write an agent artifact to the shared blackboard. Use this to persist any agent output so downstream agents can read it. |
| blackboard_readC | Read the latest artifact from the blackboard for a given run, agent, and key. |
| blackboard_listA | List all artifact keys written by a specific agent in a run. |
| scar_lookupA | Look up whether a failure pattern has been seen and resolved before. Returns the resolution if found, null otherwise. |
| scar_recordC | Record a new failure resolution in SCAR memory so it can be retrieved next time. |
| cache_getA | Check if an identical LLM request was already executed and cached. Returns the cached response or null. |
| cache_setA | Store an LLM response in the cache. Call this after receiving a response to avoid re-paying for identical future requests. |
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 7 tools
Each tool has a distinct purpose within clear subsystems (blackboard, cache, scar). There is no overlap between the three groups, and within each group, operations are differentiated by action (list/read/write, get/set, lookup/record).
All tool names follow a consistent pattern: a lowercase prefix (blackboard_, cache_, scar_) followed by a verb in snake_case. This makes it easy to infer the domain and action for each tool.
With 7 tools, the set is well-scoped for an agent coordination toolkit. Each tool serves a distinct, necessary function, and the count is neither too sparse nor overwhelming.
The toolkit covers the core operations for each subsystem: read/write/list for blackboard, get/set for cache, lookup/record for scar. Minor gaps exist, such as missing delete operations, but the essential lifecycle for agent coordination is present.