RigorPilot Skills - Setup Guide¶

Source: lllllllama/rigorpilot-skills (2.6M combined installs) GitHub: github.com/lllllllama/rigorpilot-skills Category: Research / ML Engineering First Seen: August 12, 2026 Quality Tier: 🟡 Beta (high install volume, young publisher)

RigorPilot packages the full AI-research loop as twelve agent skills: explore a research area, resolve paper context, reproduce experiments, run training, and audit the result - all with safety-first debugging. Every skill carries ~234K installs, making this one of the highest-volume research clusters on skills.sh.

Installation¶

# Install the full repo
npx skills add llllllllama/rigorpilot-skills

# Or install individually
npx skills add llllllllama/rigorpilot-skills --skill ai-research-explore
npx skills add llllllllama/rigorpilot-skills --skill paper-context-resolver
npx skills add llllllllama/rigorpilot-skills --skill safe-debug

Core Skills¶

Skill	Installs	Use For
`ai-research-explore`	235.2K	Systematic exploration of a research area
`analyze-project`	234.8K	Project structure and goals analysis
`ai-research-reproduction`	234.6K	Reproducing published AI research
`explore-code`	234.5K	Codebase exploration for research repos
`paper-context-resolver`	234.5K	Resolving paper references and context
`safe-debug`	234.5K	Debugging without breaking experiments
`repo-intake-and-plan`	234.5K	Intake a repo and produce an execution plan
`minimal-run-and-audit`	234.4K	Minimal run + result audit loop
`run-train`	234.4K	Training run execution
`env-and-assets-bootstrap`	234.4K	Environment and dataset/asset setup
`explore-run`	234.4K	Exploratory run management
`ai-paper-reproduction`	0	End-to-end paper reproduction (newest)

Prerequisites¶

Requirement	Details
Node.js + npx	For the skills.sh CLI install path
Python ML stack	PyTorch/HF ecosystem for the training-oriented skills
GPU access	For `run-train` and reproduction workloads (local or cloud)
Git	Repos under study are cloned locally

CorpusIQ Use Cases¶

Use Case	How
Model evaluation research	`ai-research-explore` + `paper-context-resolver` for surveying techniques before evals
Agent skill upgrades	`repo-intake-and-plan` + `explore-code` when assessing new tooling repos for the stack
Safe experimentation	`minimal-run-and-audit` + `safe-debug` pattern for local AI infrastructure work

Integration with CorpusIQ¶

CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.

This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.

See what CorpusIQ connects

Limitations / Verification¶

Newest skill (ai-paper-reproduction) has zero installs - treat as alpha
Verify install: npx skills list | grep -E "research|rigor" shows installed entries
Training skills assume a working GPU environment; they orchestrate, they don't provision hardware

Powered by CorpusIQ