explore-code
lllllllama/rigorpilot-skills
Auditable exploratory code modifications for deep learning research on isolated branches with rollback records.
What is explore-code?
A leaf skill for bounded, source-anchored code changes in deep learning repositories when researchers explicitly authorize isolated exploratory work. Use it to transplant modules, adapt backbones, add LoRA/adapter layers, or combine low-risk modifications while maintaining rollback-aware records separate from the trusted baseline.
- Transplant and adapt modules from source repositories into isolated branches or worktrees
- Insert LoRA or adapter layers into existing model architectures
- Replace model heads or stitch together meaningful low-risk module combinations
- Record exploratory changes with rollback instructions and scientific rationale in `explore_outputs/`
- Generate changesets, scientific changelogs, comparability reports, and run summaries
- Maintain isolation from trusted baseline and current_research to prevent contamination
How to install explore-code
npx skills add https://github.com/lllllllama/rigorpilot-skills --skill explore-code- Explicit researcher authorization for exploratory modifications
- An isolated git branch or worktree separate from the trusted baseline
- Access to `explore_outputs/` directory for recording changes and results
How to use explore-code
- 1.Confirm the researcher has explicitly authorized exploratory work on an isolated branch or worktree
- 2.Use `scripts/plan_code_changes.py` to plan the source-anchored modifications (module transplant, backbone adaptation, LoRA insertion, or head replacement)
- 3.Apply changes to the isolated branch, favoring minimal adaptation over freeform rewrites
- 4.Record the changeset, scientific rationale, and rollback instructions in `explore_outputs/CHANGESET.md`
- 5.Generate `explore_outputs/SCIENTIFIC_CHANGELOG.md` documenting why the candidate change is meaningful
- 6.Create `explore_outputs/COMPARABILITY_REPORT.md` to track how the change affects reproducibility
- 7.Use `scripts/write_outputs.py` to finalize `explore_outputs/status.json` and `explore_outputs/TOP_RUNS.md`
- 8.Hand off execution to `minimal-run-and-audit` or `run-train` if running the modified code is needed
Use cases
- Testing a backbone adaptation idea before committing to the main research pipeline
- Inserting parameter-efficient fine-tuning layers (LoRA/adapters) to evaluate their impact
- Combining modules from different sources to prototype a new architecture
- Documenting why a candidate code change is meaningful and how to revert it
- Recording exploratory runs and their results for later review without affecting verified conclusions
- Deep learning researchers conducting exploratory architecture modifications
- Research engineers prototyping low-risk module combinations
- Teams needing auditable records of candidate implementations before verification
explore-code FAQ
Use explore-code for isolated code modifications only. Use ai-research-explore when the task spans both current_research coordination and exploratory runs.
No. This skill must keep work isolated from the trusted baseline. Always work on an isolated branch or worktree.
Do not use this skill. It requires explicit researcher authorization before applying exploratory modifications.
It produces CHANGESET.md, SCIENTIFIC_CHANGELOG.md, COMPARABILITY_REPORT.md, TOP_RUNS.md, and status.json in the explore_outputs/ directory.
No. This skill is for source-anchored, low-risk modifications only. Broad refactors or from-scratch ideas are outside its scope.
Full instructions (SKILL.md)
Source of truth, from lllllllama/rigorpilot-skills.
name: explore-code
description: Rigor Improve implementation leaf skill for auditable candidate implementation in deep learning research repositories. Use when the researcher explicitly authorizes exploratory work on an isolated branch or worktree to transplant modules, adapt a backbone, add LoRA or adapter layers, replace a head, or stitch together meaningful low-risk migration ideas with rollback-aware records in explore_outputs/. Do not use for end-to-end exploration orchestration on top of current_research, trusted baseline reproduction, conservative debugging, environment setup, verified contribution claims, or default repository analysis.
explore-code
Use this as the Rigor Improve implementation leaf skill. The installed slug
remains explore-code for compatibility.
Use the shared operating principles in
../ai-research-reproduction/references/agent-operating-principles.md; this skill should guide
bounded candidate code work without over-prescribing implementation details.
When to apply
- When the researcher explicitly authorizes exploratory code changes on an isolated branch or worktree.
- When the task is source-anchored module transplant, backbone adaptation, LoRA or adapter insertion, or low-risk module combination.
- When summary-level recording is sufficient and the result is a candidate, not a trusted conclusion.
When not to apply
- When the request is for trusted baseline work, conservative debugging, or normal training execution.
- When the user did not explicitly authorize exploratory modifications.
- When the task is a broad refactor or a from-scratch idea implementation.
Clear boundaries
- This skill owns exploratory code modifications only.
- It must keep work isolated from the trusted baseline.
- Use
ai-research-exploreinstead when the task spans both current_research coordination and exploratory runs. - It may hand off execution to
minimal-run-and-auditorrun-train. - It should favor source-anchored copying and minimal adaptation over freeform rewrites.
- It should record why a candidate change is meaningful, how to roll it back, and why it remains a candidate rather than a verified contribution.
Output expectations
explore_outputs/CHANGESET.mdexplore_outputs/SCIENTIFIC_CHANGELOG.mdexplore_outputs/COMPARABILITY_REPORT.mdexplore_outputs/TOP_RUNS.mdexplore_outputs/status.json
Notes
Use references/explore-policy.md, ../ai-research-reproduction/references/research-rigor-principles.md, scripts/plan_code_changes.py, and scripts/write_outputs.py.
Related skills
More from lllllllama/rigorpilot-skills and the wider catalog.
explore-run
Plan and execute bounded exploratory runs for deep learning research with fair-comparison tracking.
minimal-run-and-audit
Execute and audit deep learning repo smoke tests with standardized evidence capture and scientific changelog tracking.
paper-context-resolver
Resolve reproduction-critical paper details when README and repo files leave gaps.
repo-intake-and-plan
README-first repository scanner for deep learning reproduction planning.
run-train
Execute and document deep learning training runs with standardized evidence capture for reproducibility.
safe-debug
Conservative diagnosis and minimal patching for deep learning training failures and errors.