mirror of
https://github.com/mudler/LocalAI.git
synced 2026-11-06 20:43:27 -05:00
Ettore Di Giacinto
2316e14632
Memory now has one switch, On or Off, in the Memory tab of the account settings. Off means nothing from the notes is put in the prompt, the model has no memory skill, tool or note kind, nothing is proposed or saved (also by triggers and other runs that start by themselves), and the model is told in one line that memory is off. The notes stay and can still be read, exported and deleted. A chat can be private: the Memory row of its Tools menu is the memory switch group, stored with the chat. The narrowest level wins: chat, agent, person, server. A saved agent cannot widen the choice of the person. The tools list says which level is off (memory.code, blocked_by, fix); the chat shows a chip and one line, and the agent form shows the person's switch as a read-only line. New API: GET and PUT /api/chats/memory for the switch and the save level, and GET /api/chats/memory/export for a zip of the notes. The Memory tab gains the switch, the three save levels, Export and Forget everything. Assisted-by: Claude:claude-sonnet-5-5
72 lines
3.8 KiB
Go
72 lines
3.8 KiB
Go
package studio
|
|
|
|
import "github.com/mudler/LocalAI/core/services/studio/builtins"
|
|
|
|
// describeGroups declares the title and the description of every switch group
|
|
// of built-in tools. The Tools menu of a chat and the editor of a saved agent
|
|
// show them next to each switch, so a person can tell what the model may do
|
|
// when it is on. A description says what the model can do, in the words of the
|
|
// tools of the group; the descriptions of the tools themselves are written for
|
|
// the model and stay as they are. A group the server does not offer is
|
|
// described all the same: the registry only lists the groups it has.
|
|
func (s *Service) describeGroups() {
|
|
for group, info := range map[string]builtins.GroupInfo{
|
|
builtins.GroupShell: {
|
|
Title: "Run commands",
|
|
Description: "The model can run shell commands and scripts. They run in an add-on that isolates them when one is installed, " +
|
|
"otherwise directly on this machine.",
|
|
WhenOff: "The model can write files but cannot run anything.",
|
|
},
|
|
builtins.GroupFiles: {
|
|
Title: "Your files",
|
|
Description: "The model can list, read, write, edit, move, delete and search your files, and read documents. " +
|
|
"A chat of an agent that works in one folder reaches only that folder.",
|
|
WhenOff: "The model cannot see or change your files.",
|
|
},
|
|
builtins.GroupWeb: {
|
|
Title: "Look things up online",
|
|
Description: "The model can search the web and read a public page by its address. Addresses inside your network are refused unless an administrator allows them.",
|
|
WhenOff: "The model cannot look anything up online.",
|
|
},
|
|
builtins.GroupMedia: {
|
|
Title: "Images and audio",
|
|
Description: "The model can describe an image or transcribe a recording from your files with a helper model.",
|
|
WhenOff: "The model cannot read images or recordings itself.",
|
|
},
|
|
builtins.GroupCollections: {
|
|
Title: "Knowledge",
|
|
Description: "The model can list your collections and search them for the passages that match a question.",
|
|
WhenOff: "The model cannot read your collections.",
|
|
},
|
|
builtins.GroupSubAgents: {
|
|
Title: "Sub-agents",
|
|
Description: "The model can start a sub-agent that works on one task with the same switches as this chat and reports back.",
|
|
WhenOff: "The model works alone.",
|
|
},
|
|
builtins.GroupSavedAgents: {
|
|
Title: "Saved agents",
|
|
Description: "The model can list your saved agents and start one as a sub-agent, with that agent's own instructions and model.",
|
|
WhenOff: "The model cannot see or start your saved agents.",
|
|
},
|
|
builtins.GroupTriggers: {
|
|
Title: "Triggers",
|
|
Description: "The model can propose, list, change and delete triggers: tasks that run by themselves at a set time or when a file arrives in a folder. " +
|
|
"A new trigger or a change waits on a card in the chat until you confirm it; pausing or resuming a trigger applies at once and you can undo it.",
|
|
WhenOff: "The model cannot create or manage triggers. A run that a trigger or a connector started never has them.",
|
|
},
|
|
builtins.GroupMemory: {
|
|
Title: "Memory",
|
|
Description: "The model can read your notes before it answers and save new ones. Turn it off to make this a private chat: nothing is read from your notes or saved to them. Your other chats are not affected.",
|
|
WhenOff: "This is a private chat. The model neither reads nor saves your notes.",
|
|
},
|
|
builtins.GroupChanges: {
|
|
Title: "Suggested changes",
|
|
Description: "The model can propose an edit to your notes, its standing rules, skills, pages and dashboard, and undo an edit you confirmed. " +
|
|
"Nothing changes until you confirm the proposal on a card in the chat.",
|
|
WhenOff: "The model cannot propose edits to your notes, its rules, skills, pages or dashboard.",
|
|
},
|
|
} {
|
|
s.builtins.DescribeGroup(group, info)
|
|
}
|
|
}
|