rea-light-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| REA_COMMAND | No | Which REA server runs behind the hub (pinned measured 139-tool release). | npx -y rea-agents@6.3.0 mcp |
| REA_TIMEOUT_MS | No | Per-call timeout; stdio serves calls one at a time. | 120000 |
| REA_PWNTOOLS_PYTHON | No | Path to a Python interpreter with pwntools 4.15.0 + pyelftools + Unicorn, required for layout inspection. | |
| REA_LIGHT_ALLOW_WRITE | No | Set to 1 to allow annotate_* / set_* write-path tools. Unset by default (read-only). |
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
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| rea_searchA | Search REA capabilities by keyword. Returns top matching tool names with one-line descriptions (no full schemas). Call rea_describe for one tool, then rea_call to run it. |
| rea_describeA | Get one REA tool's full input schema and example as markdown text. Use after rea_search, before rea_call. |
| rea_callA | Run any REA tool by name. Pass arguments as a JSON object string. Example args_json: {"document":"main","address":"0x1000"}. Get the schema via rea_describe first. |
| rea_investigateA | Investigate a target with one question (e.g. how does search work). Opens the target, runs the right REA inspection, returns markdown summary with evidence ids. Prefer this over manual rea_search for first contact. |
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 4 tools
The three pipeline tools (search → describe → call) have clearly distinct roles, and descriptions state the ordering explicitly. The only real overlap is rea_investigate versus the manual pipeline, but the descriptions resolve it by declaring investigate the preferred first-contact path.
All four tools share a uniform `rea_` prefix followed by a single verb (search, describe, call, investigate). Predictable and immediately readable as a family.
Four tools is slightly thin in absolute terms, but for a gateway/meta-server that delegates the real surface to wrapped REA tools this is the right shape — each tool earns its place in the search-describe-call loop plus one shortcut.
Discovery, schema introspection, execution, and a one-shot investigation path are all covered, with no dead ends in the core workflow. Minor gap: there is no browse/list-all operation, so an agent without a good keyword cannot enumerate the catalogue.