Skip to content
Get packs

Waypoint Microbiome Models

Embed, fine-tune, benchmark and convert microbiome abundance data with Waypoint models

Source Download skill
Repository
K-Dense-AI/scientific-agent-skills
Stars
55,780
Licence
MIT
Files
8 files

What it does

This skill teaches an AI agent to work with Outpost Bio's open microbiome foundation models (Waypoint) and the waypoint command-line tool. Microbiome samples are treated like sentences, with each taxon as a word, so the model can turn a sample into a numeric vector for clustering or classification, or be fine-tuned on labelled data. It also helps convert outputs from common profilers such as MetaPhlAn, Kraken2, QIIME 2 and MGnify into the format Waypoint expects.

Great for

  • Bioinformatics or microbiome researchers who want embeddings or a fine-tuned predictor from taxonomic abundance tables
  • Data scientists preparing MetaPhlAn, Kraken2 or QIIME 2 outputs for machine learning
  • Teams benchmarking their own microbiome model on the Compass eight-task suite

Skip it if

  • Small labelled datasets where a simple random-forest baseline with grouped validation is the better first step, since embeddings are not guaranteed to help
  • Clinical or diagnostic decisions, because the models are research artefacts and performance on a new study is not guaranteed
  • Taxa outside the model vocabulary, which are silently dropped unless you check coverage first

Try asking

  • “Convert my MetaPhlAn merged table into waypoint format and embed the samples with Waypoint-6m”
  • “Fine-tune Waypoint-45m to predict disease status from my abundance table, using grouped validation”
  • “Check how many of my taxa the Waypoint-6m vocabulary covers before I trust the results”
  • “Benchmark my model on the Compass tasks and record the protocol”
Step-by-step guide: how to use Waypoint Microbiome ModelsInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 8 original files, the MIT licence and an attribution note.

    waypoint-bio.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip waypoint-bio.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip waypoint-bio.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Use full lineage strings (k__...;g__...) rather than bare species names, so the tokenizer can fall back to higher ranks when genus is missing.
  • Run the vocab coverage script before training or benchmarking to see how much abundance your model can actually use.
  • Pin a reviewed Hugging Face commit and read the upstream review notes before training, because the tokenizer loader runs repository code and some training features are unfinished.

What's inside

  • references/cli-reference.md
  • references/compass-benchmark.md
  • references/data-preparation.md
  • references/python-api.md
  • references/upstream-review.md
  • scripts/profiler_to_waypoint.py
  • scripts/vocab_coverage.py
  • SKILL.md

Redistributed unmodified from K-Dense-AI/scientific-agent-skills under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.