← all skills

tomkaltofen

tomkaltofen

@tomkaltofen

0 followers · 2 skills

2 published skills

  1. Mloda · tomkaltofen
    Give an AI agent declarative, deterministic data access via mloda (https://github.com/mloda-ai/mloda) - a Python plugin framework where the caller describes WHAT features or data it needs and mloda resolves HOW to compute or fetch them across a plugin graph, with built-in lineage back to source. Use when an agent needs to check whether an existing mloda plugin already covers a requested feature (see the mloda-plugins skill for writing a new one), before requesting structured data or feature aggregations as a tool call instead of writing ad hoc fetch/transform code (the "LLM Tool Function" pattern), assembling a multi-source context window declaratively, or chaining retrieval/validation/redaction steps for a RAG pipeline by feature name.
    4
    installs
  2. Mloda Plugins · tomkaltofen
    Guide an AI agent through building mloda (https://github.com/mloda-ai/mloda) plugins: FeatureGroup, ComputeFramework, and Extender classes. Use to check the mloda-registry index for an existing plugin before writing one, when writing or reviewing a FeatureGroup/ComputeFramework/Extender implementation, or when packaging, sharing, or publishing a plugin to mloda-registry.
    4
    installs