Incorporating generative AI can greatly boost productivity in knowledge work, with a clear call for developers to provide the approaches that mitigate pitfalls and maximize ergonomics¹:

Developers can also help with figuring out where AI can be inserted into workflows and how to design technology for doing that.

Originally, this explored building an agent framework from scratch and how to design agents for specific Synapse workflows -- from basic things like prompts and tools needed to potentially more useful/advanced framework features. For example, interesting things included integrating symbolic AI to work in tandem with generative AI or smart delegation between different LLM providers.

For practical usage, this is nowadays used much more often as a local MCP server with agentic client/frameworks like Claude Desktop/Code and especially the open-source goose, which have all the real, recent developments one would wish for.

Science is built upon data. Research communities have the goal of sharing data packaged and disseminated optimally for (re)use.

If the community is lucky, this is supported by dedicated data curators/managers or contributing scientists directly involved as curators.

If the community is even luckier, AI agents will assist human curators to help with curation workflows, such as harmonizing data, finding data, assessing data, etc.

Several options are available:

1. accent can be run as an MCP server so you can use your favorite MCP-compatible agentic client interface. See MCP Server mode. It might also be helpful to read doc/MCP.md.
2. If you don't have a preferred compatible client, accent does include a basic terminal interface as well as a web app client interface to test our agents. To use either of these, see Built-in client interface. But these were only for very early concept demos and development of client UI/UX has not been prioritized since.

accent can be run as a local MCP server that you use with your preferred desktop client.

1. Download a jar release v0.6.0 or later from the releases page.
2. Follow the configuration instructions for your client. With Claude for Desktop, use the configuration below.

{ 
  "mcpServers": {
    "accent": {
      "command": "java",
      "args": [
        "-Dclojure.tools.logging.factory=clojure.tools.logging.impl/log4j2-factory",
        "-Dorg.eclipse.jetty.util.log.class=org.eclipse.jetty.util.log.Slf4jLog",
        "-Dlog4j2.contextSelector=org.apache.logging.log4j.core.async.AsyncLoggerContextSelector",
        "-Dlog4j2.configurationFile=log4j2-mcp.xml",
        "-jar",
        "/Users/your/path/to/accent.jar",
        "mcp-server"
      ],
      "env": {  
        "SYNAPSE_AUTH_TOKEN": "your-synapse-pat"
      }
    },
    ...
  }
}

You can use the built-in interface in accent, though it's rather minimal.

1. For the web app client interface, download a jar release v0.4.0 or later from the releases page.
2. Set up a config file in the same location as your jar. See Configuration.
3. Choose whether you want to run the web app or console UI below:

This is recommended for most users. Run the jar by double-clicking the file. This should open a browser window with your default browser app with Syndi as your assistant.

This requires Clojure dev tooling and is only recommended if you're comfortable with Clojure (or want to be comfortable with Clojure) and want to test experimental features and hack with the interface/framework internals. It does give more control for running specific agents that might not be available in the web app.

1. Install Clojure.
2. Clone this repo.
3. Set up a config file at the root of the repo. See Configuration.
4. Run the desired agent module, e.g. clj -M -m agents.syndi.

Settings and (optionally) credentials ares defined in config.edn. Review the example_config.edn file; rename it to config.edn and modify as needed.

The app integrates two providers, Anthropic and OpenAI, and an initial model provider must be specified. In the same chat, it is possible to switch between models from the same provider but not between different providers, e.g. switching from ChatGPT-3.5 to ChatGPT-4o, but not from ChatGTP-3.5 to Claude Sonnet-3.5. However, existence of the switching feature does not suggest that the user should be manually and frequently switching between models. For both providers, the default is to use a model on the smarter end, though later on it may be possible to specify an initial model in the config. Tip for usage: Trying to reduce costs by switching to a cheaper model for some tasks is likely premature optimization at this early stage.

• OpenAI
  • To use, must have OPENAI_API_KEY in env or set in config.
  • The default model is ChatGPT-4o.
• Anthropic
  • To use, must have ANTHROPIC_API_KEY in env or set in config.
  • The default model is Claude Sonnet 3.5.

• doc/AGENTS.md describes and explains more about agents and usage.
• doc/MCP.md describes and explains more about Model Context Protocol.
• doc/ROADMAP.md describes roadmap and context.
• doc/DESIGN.gv reflects early draft architecture of app.

Planned demo materials will be linked once available:

• Assisted curation workflow for preparing some kind of data asset for Synapse (e.g. a dataset).
• Data model exploration and development for working with different DCC-specific models to reuse concepts, maintain alignment, improve quality, etc.

Nothing more is planned until after the Evaluation (below).

Feedback is currently being gathered with curators who are being trained for integrating this into their workflows. The comparisons will be between workflows that:

1. Doesn't incorporate any LLM and does things manually, maybe with custom scripting, or with some other non-AI app.
2. Incorporates generative AI but only via the default online chat interface.
3. Incorporates generative AI through a different custom interface/solution.