Skills lllllllama rigorpilot-skills RigorPilot
run-train
- Use it when:
- A training command is already chosen and you want it started, checked over a short run, kicked off in full, or resumed.
- It does:
- Runs the command through RigorPilot's shared runtime and records the command, config, seed, logs, checkpoints and metrics.
- You get:
- A train_outputs/ folder: summary, commands, log, status, scientific changelog and comparability report, with partial, blocked and resumed states kept apart.
- Won't:
- Run sweeps, change the architecture, or present a startup check as a reproduced result.
- Usually called by:
- RigorPilot's main skill, ai-research-reproduction, for its training steps.
Skill 7 of 11 in the RigorPilot bundle, after minimal-run-and-audit and before safe-debug.
Only works inside its bundle: Its run and report scripts load a shared runtime that ships inside ai-research-reproduction, the repo's main skill, and stop with an error without it. Install RigorPilot instead.
TL;DR
The training step of RigorPilot, a set of skills that reproduces deep-learning research code from its README; minimal-run-and-audit covers inference and evaluation instead. You give it the goal, the command, what the environment and data are assumed to be, and the mode: startup verification, a short verification run, a full kickoff, or a resume. It records the exact command and the data and checkpoint assumptions and, when the logs show them, the steps completed, best metric, best checkpoint and why the run stopped, without inventing values when logs are silent. Long runs are watched for a limited window rather than indefinitely. Conclusions stay cautious for short or partial runs, and the scientific changelog and comparability report say what could change the result's meaning. Planning experiments, choosing data subsets and comparing variants are left to the main skill or the exploration skills.
Summarizes the skill at fb3ccdf.
npx skills add https://github.com/lllllllama/rigorpilot-skills --skill run-train skills.sh pageSource on GitHubAll skills from lllllllama/rigorpilot-skills
More weekly downloads than 93% of skills
Weekly values
| Week to | Installs |
|---|---|
| Sep 21 | 17 |
| Sep 28 | 22 |
| Oct 5 | 25 |
| Oct 12 | 3,487 |
| Oct 19 | 10,902 |
| Oct 26 | 11,030 |
| Nov 2 | 8,738 |
| Nov 9 | 7,264 |
More downloads than 94% of skills
Fewer stars than 88% of repos
Newer than 97% of skills
| Category |
Research reproduction, financial data, charts and diagrams. |
|---|---|
| Type |
A multi-step process the agent follows: interview, debug, run a TDD loop, review. |
| Effect |
Its own steps create or change files, git state or local tools. |
| Needs |
Instructions only. Works in any agent that supports skills. |
| Publisher |
Everyone else. Unverified vendor claims default to community. |
| Topics | |
| Stack |
SKILL.md
From lllllllama/rigorpilot-skills at fb3ccdf. Checked daily, last changed Nov 6, 2026.
Frontmatter
name: run-train
description: Rigor Train skill for deep learning research repositories. Use when a documented or selected training command should be run conservatively for startup verification, short-run verification, full kickoff, or resume, with command, config, seed, log, checkpoint, status, and metric evidence written to standardized `train_outputs/`. Do not use for environment setup, exploratory sweeps, speculative idea implementation, or end-to-end orchestration.run-train
Use this as the Rigor Train skill. The installed slug remains run-train for
compatibility.
Use the shared operating principles in
../ai-research-reproduction/references/agent-operating-principles.md; this skill should keep
training evidence bounded while leaving repository-specific monitoring details
to the model.
When to apply
- When the training command has already been selected and should be executed conservatively.
- When the researcher wants startup verification, short-run verification, full training kickoff, or resume handling.
- When the run needs structured training status, checkpoint, and metric reporting.
When not to apply
- When the main task is environment setup or asset download.
- When the researcher wants inference-only or evaluation-only execution.
- When the task is speculative exploration, multi-variant sweeps, or autonomous idea implementation.
- When the user still needs repository intake or paper gap resolution.
Clear boundaries
- This skill executes a selected training command and normalizes the resulting evidence.
- It does not choose the overall research goal on its own.
- It does not own exploratory branching or speculative code adaptation.
- It should record partial, blocked, resumed, and kicked-off states clearly.
- It should preserve reproducibility context such as configs, seeds, checkpoints, logs, metrics, and runtime assumptions when available.
Input expectations
- selected training goal
- runnable training command
- environment and asset assumptions
- run mode such as startup verification, short-run verification, full kickoff, or resume
Output expectations
train_outputs/SUMMARY.mdtrain_outputs/COMMANDS.mdtrain_outputs/LOG.mdtrain_outputs/SCIENTIFIC_CHANGELOG.mdtrain_outputs/COMPARABILITY_REPORT.mdtrain_outputs/status.json
Notes
Use references/training-policy.md, ../ai-research-reproduction/references/deep-learning-experiment-principles.md, scripts/run_training.py, and scripts/write_outputs.py.