Runs on your machine AGPL-3.0 GitHub

Screenshots

Tzara, running locally. Click to view.

The wiki

  • A rendered wiki page.
    Markdown document served as a formatted webpage with code and LaTeX. Ocean theme.
    A rendered wiki page.
  • A rendered wiki page.
    Markdown document served as a formatted webpage with LaTeX, and visible side bar with a table of contents, tags, and summary. Ocean theme.
    A rendered wiki page.
  • Dark mode view of a rendered page.
    Dark mode and light mode follow browser settings automatically. Ocean theme.
    Dark mode view of a rendered page.
  • Light mode view of embedded code.
    Default theme showing embedded code with inline execution buttons.
    Light mode view of embedded code.
  • Light mode view of embedded code with widgets.
    Default theme showing embedded code execution results with widgets.
    Light mode view of embedded code with widgets.
  • Light mode view of embedded code with widgets, sage color scheme.
    Sage theme showing embedded code execution results with widgets and a Plotly graph of retrieved stock data.
    Light mode view of embedded code with widgets, sage color scheme.

Management Pages

  • Jupyter kernel management page.
    Each page gets its own isolated kernel that gets reaped when inactive.
    Jupyter kernel management page.
  • A rendered wiki page.
    Agent activity page. Any staged edits appear here for acceptance or refusal.
    A rendered wiki page.
  • Model mangment page with model details.
    Model management page allows for downloading and switching chat models. An embedding model change is a rebuild and reindex.
    Model mangment page with model details.
  • Monitoring dashboard for application health.
    A monitoring page surfacing any errors or issues that may occur. In this page an agent is failing because the vault it is configured for does not exist.
    Monitoring dashboard for application health.
  • Agent listing page for manual runs.
    A task managment page to look at existing and past background tasks, as well as for manually running agent tasks.
    Agent listing page for manual runs.
  • .
    Sage theme showing embedded code execution results with widgets and a Plotly graph of retrieved stock data.
    .

Editing

  • Editing a document with popup menu.
    Editing is done with Codemirror 6. Pressing "/" after a space or with selected text brings up a tool menu. You can type the partial command, such as "/list" and the menu filters. Files can be added to the wiki with the drag and drop area. A link to the file is added in the editor.
    Editing a document with popup menu.
  • Document modified by chatting need human approval.
    If you ask for a section to be added or any modifications at all, changes are staged on a shadow copy of the document. An approval gate exists because small langauge models can be anti-cromulent. And I also don't want to remove the agency of the human.
    Document modified by chatting need human approval.
  • Document modified by chatting.
    The document updates immediately with changes when accepting suggested edits from the document agent.
    Document modified by chatting.
  • Chats are private.
    Chats are private and are not logged. If you leave a page or refresh, the record of the chat is gone, however, your chat session (which will eventually time out) might still have your conversation in it. Hit the Reset button to clear it.
    Chats are private.

Chatting

  • Chatting with the whole vault.
    Vault wide chatting has a different tool set as well as a different system prompt than document chat. Here there are batched document creation tools available.
    Chatting with the whole vault.
  • Document modified by chatting need human approval.
    If you ask for a section to be added or any modifications at all, changes are staged on a shadow copy of the document. An approval gate exists because small langauge models can be anti-cromulent. And I also don't want to remove the agency of the human.
    Document modified by chatting need human approval.
  • Document modified by chatting.
    The document updates immediately with changes when accepting suggested edits from the document agent.
    Document modified by chatting.
  • Chats are private.
    Chats are private and are not logged. If you leave a page or refresh, the record of the chat is gone, however, your chat session (which will eventually time out) might still have your conversation in it. Hit the Reset button to clear it.
    Chats are private.

More Pages

  • Index view of files in the vault.
    The Index page shows all files in the vault. You can drag and drop files to move them. Any links to the files are automatically fixed, this is configurable of course.
    Index view of files in the vault.
  • Work in progress of a graph view of a vault.
    A graph view of a vault using the .canvas format. Work in progress.
    Work in progress of a graph view of a vault.
  • A view of a canvas file embedded in a page.
    Canvas files (as well as images, pdfs, csv, and markdown pages) can be embedded in other pages.
    A view of a canvas file embedded in a page.
  • List of git actions on a page.
    Each page is version controlled with git, so we can see a history when each file is changed.
    List of git actions on a page.
  • Git style diffs on a page.
    Diff comparison's with the current version of the document.
    Git style diffs on a page.

Agents

  • An agent memory file.
    Agents can have optional memory, which is a capability you grant when creating it. You can specify a custom memory prompt for the agent or use Tzara default. Memory is a two step process that happens after an agent completes its task or hits a user defined iteration limit, where one step is to write a memory file, and an optional second step to write to a ledger file. Memory is fuzzy, Ledgers are not.
    An agent memory file.
  • Agent log activity listing.
    Agents can do some interesting things. Sometimes those things are stupid. Create a good prompt for an agent is not easy. Different LLM models can affect quality, but your prompts are really the foundation for them. If you don't know what an agent will do, you can enable loggings for an agent. There is limit or cleanup of logs. That's a human process.
    Agent log activity listing.
  • An agent log file.
    Agent log files keep a record of lots of activity and aspects of the run.
    An agent log file.