backskills.ai Blog · llms.txt · agents.md · Most used by agents · Installs · Suggest · Monitor · Submit your skill design-experiment github:niznik-dev/cruijff_kit · first committed 2025-10-22 · source · badge
Plan LLM fine-tuning and evaluation experiments. Use when the user wants to design a new experiment, plan training runs, or create an experiment_summary.yaml file.
Skill Health: not yet scored Listed, not yet scored. This skill has no public model test yet. It is in the scoring queue; submitted skills go first, then the most-installed.
Version history 41 commits to this file on GitHub (newest first).
2026-06-15 · commit
2c30ea9 · Teach design-experiment the prompt-sweep recipe
2026-06-15 · commit
9d9629c · Address Slate/Sage review of the system_prompt collapse
2026-06-11 · commit
93ac849 · Home dataset_type on controls.dataset_type, not data.training (#478)
2026-06-11 · commit
bb3b00b · Decouple eval-only scaffolding from torchtune (#478)
2026-05-19 · commit
206899b · Tighten preflight blurb in all four skills
2026-05-19 · commit
8e65437 · Add env preflight to scaffold/run/analyze SKILL.md
2026-05-19 · commit
cc62e55 · Add design-experiment preflight env check
2026-05-07 · commit
eab9350 · Rename blueprint dir + drop legacy 'synthetic'/'sanity check' framing
2026-05-06 · commit
88ad8fc · Iterate on /ck-setup based on PR review
2026-05-05 · commit
188d142 · Add /setup skill and trim design-experiment prereq inline check
2026-04-28 · commit
dcbddaa · Skill audit batch 3: nice-to-have cleanups
2026-04-05 · commit
cb47f53 · updated skills workflow
2026-02-27 · commit
a9c4634 · Standardize experiment artifact locations (#325)
2026-02-06 · commit
686da47 · Merge pull request #287 from niznik-dev/283-analyze-experiment-report
2026-02-06 · commit
a9a6b08 · Convert design-experiment logging and add Mistral tokenizer (#292)
2026-02-05 · commit
53cade4 · Make placeholder validation required, not optional
2026-02-05 · commit
660284c · Add claude.local.md validation and clean up template
2026-02-05 · commit
801bcef · Add report generation to analyze-experiment skill
2026-01-05 · commit
700f729 · Theme 9: Clarify sanity check vs workflow test terminology
2026-01-05 · commit
54b38e6 · Merge main into pre-yaml-refactor
2025-12-24 · commit
0a36d10 · Use config_path pattern for train/eval parity
2025-12-16 · commit
3b40159 · slight cleanup before merge
2025-11-28 · commit
8b80c4f · refactor design-experiment to generate standard yaml and use jsonl for machine-readable logging
2025-11-14 · commit
cee7a70 · Merge Sarah's agent-based scaffolding architecture
2025-11-11 · commit
c8e12c3 · Refactor design-experiment to use action verb pattern
2025-11-11 · commit
d319c9c · move scaffold* tasks to agents
2025-11-11 · commit
b75db17 · Modularize design-experiment skill with components + validation pattern
2025-10-30 · commit
ac55310 · Document 0-indexed epoch convention in architecture and skills
2025-10-29 · commit
6d08320 · Update skill instructions for new data location
2025-10-29 · commit
11fdd80 · Merge main into 176-refactor-design-experiment-skill
2025-10-29 · commit
a64137a · Replace hard-coded paths with placeholders for portability
2025-10-29 · commit
046427f · Add template examples for scaffold skill compatibility
2025-10-29 · commit
8a0153c · Add Configuration section example with explicit system prompt
2025-10-29 · commit
9a06c04 · Remove redundant log format template section
2025-10-29 · commit
d4c1ece · Replace runtime commands with design-phase commands
2025-10-28 · commit
df30d8e · Refactor design-experiment skill per issue #176
2025-10-28 · commit
b5a9364 · Rename probes/ to sanity_checks/ throughout codebase
2025-10-27 · commit
9428836 · Refactor skills to support multiple preparation and evaluation tools
2025-10-24 · commit
61ce38e · Rename tasks/ to experiments/ to eliminate terminology confusion
2025-10-24 · commit
583590b · Refactor skills into modular orchestrator/worker architecture
2025-10-22 · commit
6f7bd44 · Rename design_experiment skill to design-experiment
Archived snapshots