Transmute

SkillDev tools

Transmute papers into experiments — connect literature to research directions

Available today. Use it from your connected AI after setup.

No other account needed.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the Transmute skill

What this skill tells your AI

The instructions your AI receives, as published by rlacombe/distillate in .claude/skills/transmute/SKILL.md and read by ahel’s review.

Turn base knowledge (papers) into gold (experiment ideas). Bridge the library to the lab.

The user provides a research question, experiment name, or topic area.

Search the library: Call mcp__distillate__search_papers with relevant keywords
Deep read: For the top 3-5 papers, call mcp__distillate__get_paper_details to read highlights, notes, and key findings
Cross-pollinate: Call mcp__distillate__synthesize_across_papers to find common techniques, disagreements, and open questions
Bridge to experiments: If a target experiment exists, call mcp__distillate__suggest_from_literature for concrete steering ideas grounded in the papers
Extract baselines: Call mcp__distillate__extract_baselines to pull quantitative targets from the papers
Propose transmutations: Synthesize 3 concrete experiment modifications:
- Each references the specific paper(s) that inspired it
- Include expected impact and confidence level
- Suggest specific hyperparameters or architectural changes
Link: Call mcp__distillate__link_paper to connect relevant papers to the experiment for future reference

Advanced