askelelem

Version, currently main branch1 version
  • main branchlatestJul 13, 2026

github.com/nogginly/askelelem.cr

A way to ask an LLM to perform actions from a Crystal application

0 stars
0 dependents
License: MPL2

Installation

# Add this to your shard.yml
dependencies:
  askelelem:
    github: nogginly/askelelem.cr
    branch: main

main is a branch, not a release, so this tracks it rather than pinning a version.

Then run:

shards install

shard.yml

Crystal
no constraint declared
License
MPL2
Author
nogginly
Targets
  • test01 from samples/test01.cr
  • test02 from samples/test02.cr
  • test03 from samples/test03.cr
  • test04 from samples/test04.cr
  • test05 from samples/test05.cr

Dependencies

Runtime Dependencies

Development Dependencies

  • ameba*github: crystal-ameba/amebadev
  • wiretap*github: nogginly/wiretap.crdev
  • spectator*gitlab: arctic-fox/spectatordev

README

AskLLM

An API shard to ask an LLM to perform actions from a Crystal application

|WARNING | |:------------------:| |Very experimental.| |Beware. |

Rationale

As a start I want to be able to setup and make single requests to an LLM. The application is responsible for all the state / session. The goal is to be able to make different kinds of calls:

  • Get an embedding vector
  • Ask a question, get a response
  • Ask a question with a JSON response schema, get a JSON response
  • Ask a question with accompanying tools, handle any tool-calling turns and get the final response

Installation

  1. Add the dependency to your shard.yml:
dependencies:
  askelelem:
    github: nogginly/askelelem.cr

Usage

Coming soon

Development

See DEVELOPMENT for how to build askelelem and run the samples.

Contributions, by invitation!

With apologies, at this time contributions are by invitation only and limited to people I know and see often.

These are early days for AskElelem and I am busy with family and work.

At this time I want to work on this at a manageable pace.