Skip to main content

Agents to support fractale agentic framework

Project description

fractale agents

Agents to use with fractale, agentic state-machine orchestrator for Science

PyPI version DOI

https://github.com/converged-computing/fractale/raw/main/img/fractale-small.png

A sub-agent is a single step in a state machine that can call other sub-agents, tools, or prompts, and respond dynamically to work on a scoped task. Sub-agents are created in order to map our expertise and logic into a more controlled execution or interaction with an LLM.

Agents

The following agents are availble.

Name Description Path
flux-operator expert to deploy Flux Operator MiniClusters to Kubernetes fractale_agents.kubernetes.FluxOperatorAgent
flux-build optimized to build containers for the Flux Operator fractale_agents.kubernetes.FluxBuildAgent
result_parse Parse specific metrics from output logs fractale_agents.parsers.ResultParserAgent
optimize General optimization agent fractale_agents.optimize.OptimizeAgent
job-transform Job specification transformation agent fractale_agents.hpc.job.JobTransformAgent
job-generate Job specification generation agent fractale_agents.hpc.job.JobGenerationAgent

The general prompt agent is provisioned by fractale directly, fractale.agents.general.PromptAgent. Would you like to see an expert added? Please open an issue and let us know.

Usage

A sub-agent is a specialized expert that can be imported and used in a fractale workflow. You can register agents on the fly, or via a configuration file.

# Register a sub-agent tool on the fly
fractale prompt -t fractale_agents.kubernetes.FluxOperatorAgent Deploy lammps to Kubernetes using the Flux Operator

Or write the tool into a registry file:

# sub-agents.yaml
tools:
  - path: fractale_agents.kubernetes.FluxOperatorAgent

And then:

fractale prompt -r ./sub-agents.yaml Deploy lammps to Kubernetes using the Flux Operator

The agents have logical dependencies. For example, the flux-operator sub agent is expecting to have access to a Kubernetes cluster to install and use the Flux Operator!

License

HPCIC DevTools is distributed under the terms of the MIT license. All new contributions must be made under this license.

See LICENSE, COPYRIGHT, and NOTICE for details.

SPDX-License-Identifier: (MIT)

LLNL-CODE- 842614

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

fractale_agents-0.0.13.tar.gz (33.9 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

fractale_agents-0.0.13-py3-none-any.whl (42.2 kB view details)

Uploaded Python 3

File details

Details for the file fractale_agents-0.0.13.tar.gz.

File metadata

  • Download URL: fractale_agents-0.0.13.tar.gz
  • Upload date:
  • Size: 33.9 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.2

File hashes

Hashes for fractale_agents-0.0.13.tar.gz
Algorithm Hash digest
SHA256 3e051c68e7a227036856e1fff7bd170b39443ca9a275f21d266267a48e98bbf0
MD5 bd9b85877cfebea2af5b006569a2de4c
BLAKE2b-256 b3fa61be013ac2163b757b605f4dc439675a0a61c76d6caeeaf3a150f1118c1d

See more details on using hashes here.

File details

Details for the file fractale_agents-0.0.13-py3-none-any.whl.

File metadata

File hashes

Hashes for fractale_agents-0.0.13-py3-none-any.whl
Algorithm Hash digest
SHA256 0e56a7d972e4df43bfaf47463a6d0eb4f26113e13f619e922f7890474ba1d90d
MD5 570c08a224dd067a7bf4488987075287
BLAKE2b-256 816a3c34eb46b54baf36f02240ebdfa87b912c3b2bd164df17d48954695e1691

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page