Description
explore-code is an agent skill from the rigorpilot-skills collection for making bounded, exploratory code changes to a deep learning research repository. It is the Rigor Improve leaf skill, and its installed name is kept for compatibility. The skill is meant for the moments when a researcher explicitly authorizes code changes on an isolated branch or worktree, so experiments do not touch the main code. It fits source-anchored tasks such as transplanting a module from a reference implementation, adapting a backbone, inserting LoRA or adapter layers, and combining low-risk modules. The result of such a change is a candidate, not a trusted conclusion, so the skill is set up to record a summary of the work rather than claim success. It guides candidate code work without over-prescribing how the implementation should look, which leaves the model's judgment about the code to work with the specific repository. It is for researchers who want an assistant to try a modification quickly and keep a clear trail of what was changed. It should not be used for ordinary bug fixes or for work that has not been authorized. On skills.sh, Socket and Snyk pass, and Gen Agent Trust Hub shows a warning. Install it with npx skills add https://github.com/lllllllama/rigorpilot-skills --skill explore-code.…
Highlights
- Bounded exploratory code changes on an isolated branch or worktree
- Suits module transplant, backbone adaptation and LoRA or adapter insertion
- Records a candidate result, not a trusted conclusion
- Requires the researcher to authorize the change
Ratings & Reviews
No ratings yet
Be the first to rate explore-code.
People also visit
explore-code FAQ
When should I use it?
When the researcher explicitly authorizes exploratory code changes on an isolated branch or worktree.
What kinds of change does it fit?
Source-anchored module transplant, backbone adaptation, LoRA or adapter insertion, and low-risk module combination.
Is the result proven?
No. The result is a candidate, and the skill records a summary rather than a trusted conclusion.
How do I install it?
Run npx skills add https://github.com/lllllllama/rigorpilot-skills --skill explore-code.
Which security audits does it have?
On skills.sh, Gen Agent Trust Hub shows a warning, and Socket and Snyk pass.
What are the best explore-code alternatives?
Popular alternatives to explore-code include analyze-project, ai-research-reproduction and explore-run. See all explore-code alternatives