doubao-jev-agent
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| JEV_API_KEY | No | Your own JEV API key from TypeSafe. If not set, the server runs in mock mode. |
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": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| jev_decideD | Make a constrained decision with the JEV Decision Engine. |
| agent_runC | Run JEV skill selection followed by local Skill Executor execution. |
| list_skillsA | List skill names supported by this PermitMCP instance. |
| controlled_agent_runC | Run proposed local actions through policy, JEV, permits, and sandbox tools. |
| approve_actionC | Approve and resume one action currently waiting for caller review. |
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 5 tools
list_skills and approve_action are clearly distinct, while jev_decide, agent_run, and controlled_agent_run overlap around JEV execution. The descriptions do differentiate the latter three by scope, but an agent must read carefully to choose between plain and controlled runs.
All names use snake_case lower_case, but the structural pattern is mixed: list_skills and approve_action are verb_noun, jev_decide and agent_run invert to noun_verb, and controlled_agent_run adds an adjective. Readable, but not a predictable convention.
Five tools is well-scoped for a JEV/PermitMCP agent server. Each tool maps to a distinct phase: list, decide, run, controlled run, and approve.
The core run-and-approve lifecycle is present, but obvious operations are missing: no reject/deny action, no list pending approvals, no action status query, and no skill detail lookup. These gaps create dead ends in the approval workflow.