web-gpt
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| WEB_GPT_DATA_DIR | No | Optional directory to override the default local data directory (default: ~/Library/Application Support/codex-web-gpt/) |
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 |
|---|---|
| ask_chatgptA | Send one task and wait. Choose effort: normally xhigh; instant/medium/high for simpler faster tasks; pro for very complex tasks. Chrome starts with saved login. Omit session_id for a new chat; reuse it to restore/continue. project_url overrides the default project for new chats. timeout: 1–840 seconds. |
| read_chatgptA | Restore and read a previous chat without sending a message. Chrome starts automatically with saved login. timeout: 0–840 seconds; use 0 for a snapshot after loading. |
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 2 tools
ask_chatgpt sends tasks and awaits responses, while read_chatgpt only restores and reads previous chats without sending. Their purposes are distinct and unlikely to be confused.
Both tools follow a consistent verb_chatgpt snake_case pattern: ask_chatgpt and read_chatgpt. The naming convention is predictable and clear.
With only 2 tools, the surface feels thin even though both tools are clearly useful. The server covers a narrow scope, but the count is at the low end of the acceptable range.
The core workflow of sending a task and reading a previous chat is well covered. Minor gaps exist, such as no explicit delete or list-chats operation, but agents can work around them using session_id handling.