@lllllllama

minimal-run-and-audit captures execution evidence and normalizes outputs for smoke tests and short verification runs in deep learning repos, with mandatory scientific changelog and comparability reporting to preserve audit trails.

run-train executes a pre-selected training command in deep learning repos, normalizing outputs into structured evidence artifacts (logs, metrics, checkpoints, comparability reports) while preserving reproducibility context and run state.

minimal-run-and-audit captures execution evidence and normalizes outputs for smoke tests and short verification runs in deep learning repos. intended for use after targets and setup are already decided, producing standardized reports on scientific comparability and result status.

env-and-assets-bootstrap prepares conda environments and asset paths for deep learning repo reproduction, translating readme setup steps into conservative environment assumptions before execution.

env-and-assets-bootstrap prepares conda environments and asset paths for deep learning repo reproduction, translating readme setup steps into conservative environment assumptions before execution.

explore-run plans bounded exploratory deep learning runs with candidate ranking by cost, success rate, and expected gain. outputs ranked candidates and comparability reports rather than certified results.

analyze-project provides read-only exploration of deep learning repositories by mapping model structure, training and inference entrypoints, config relationships, and patterns. guidance is conservative, non-prescriptive.

ai-research-explore is a rigor-focused skill for deep learning research exploration after explicit researcher authorization, emphasizing reproducibility and scientific standards over autonomous discovery claims.

Main orchestrator for README-first AI repo reproduction. Use when the user wants an end-to-end, minimal-trustworthy reproduction flow that reads the repository…

Rigor Debug / Rigor Audit skill for deep learning research work. Use when the user pastes a traceback, terminal error, CUDA OOM, checkpoint load failure, shape…

Rigor Improve implementation leaf skill for auditable candidate implementation in deep learning research repositories. Use when the researcher explicitly…

Rigor Reproduce compatible skill slug for README-first deep learning repository reproduction. Use when the user wants an end-to-end, minimal-trustworthy flow…

Rigor Intake helper for README-first deep learning repo reproduction. Use when the task is specifically to scan a repository, read the README and common…

Rigor Paper Context helper for README-first deep learning repo reproduction. Use only when the README and repository files leave a narrow reproduction-critical…

Main orchestrator for README-first AI repo reproduction. Use when the user wants an end-to-end, minimal-trustworthy reproduction flow that reads the repository…

Rigor Intake helper for README-first deep learning repo reproduction. Use when the task is specifically to scan a repository, read the README and common…

Rigor Paper Context helper for README-first deep learning repo reproduction. Use only when the README and repository files leave a narrow reproduction-critical…