@perssua/mcp
OfficialServer Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| PERSSUA_LAUNCH_URL | No | Hosted launcher page (e.g. https://perssua.com/launch) used by create_session_link to produce https links that chat UIs reliably render. | |
| PERSSUA_MCP_SOURCE | No | Default source tag stamped on sessions (claude, chatgpt, grok, …). | |
| PERSSUA_USER_DATA_DIR | No | Override the app's user-data directory (defaults to the bridge file, then platform defaults). |
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
} |
| prompts | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| app_statusA | Check whether the Perssua desktop app is installed and running on this machine, and where its integration bridge lives. |
| list_assistantsA | List the user's configured Perssua assistants (name and id) so a session can be started with the right one. |
| start_sessionA | Launch the Perssua desktop app and start a session with an optional assistant, an initial prompt, free-text context, and text files attached as context. Runs on the same machine as the Perssua app; for remote/hosted setups use create_session_link instead. Files must be paths to local text files (binary files are skipped). To start with a NEW assistant that does not exist yet, use create_assistant instead. |
| create_assistantA | Create a new custom assistant in the Perssua desktop app (name + system-prompt instructions, optional knowledge) and open a session with it. BEFORE calling this, interview the user briefly so the assistant fits: (1) what is the assistant's goal / what sessions will it support, (2) how should it respond (tone, format, language), (3) what knowledge should it carry (notes, files, background), (4) what should the first session start with. Then write the instructions yourself from those answers. Knowledge text and files become the assistant's permanent context, not part of the first message. Runs on the same machine as the Perssua app. |
| create_session_linkA | Build a perssua:// deep link (and, when configured, an https launcher link) that starts a Perssua session with an assistant, prompt, and context when the user clicks it. Use this from hosted/remote connectors (ChatGPT, Grok, web chats) where this server cannot reach the user's machine. Inline links never auto-submit — the user reviews the prefilled prompt in Perssua. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| new_assistant | Interview the user about the assistant they need, then create it in Perssua and start the first session. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 5 tools
Each tool maps to a distinct action: environment status, assistant listing, remote link creation, local session launch, and assistant creation. The two session-starting tools are clearly separated by local vs remote context and explicitly reference each other, reducing ambiguity.
Most tool names follow a clear verb_noun snake_case pattern: list_assistants, create_session_link, start_session, create_assistant. app_status is the only noun-phrase name without an action verb, a minor deviation from the otherwise consistent convention.
Five tools is well-scoped for a desktop-app integration bridge. Each tool covers a distinct step in the assistant/session workflow without unnecessary overlap or bloat.
The set covers the core workflow: checking app availability, listing assistants, starting sessions locally or via links, and creating new assistants. It lacks update/delete or session-management operations, but those may realistically live in the desktop app itself.