Menu
Skills in this repository
SkillsMP has collected 90 skills from SamyakJhaveri/loam. Open a skill to review its source and details.
SamyakJhaveri/loamShowing 10 of 90 collected skills.
Adversarial research interrogation before launching an eval or making a published claim. Use before /overnight-eval, before adding a claim to a paper, or when a reviewer-style stress test is needed. 4 waves: basics, clarifications, deep probes, assessment.…
Code review checklist for parallel code correctness in CUDA, OpenMP, and OpenCL. Use when reviewing LLM-generated parallel code translations, when auditing benchmark source code for correctness issues, when writing the paper's 'common failure modes' section,…
Structured hypothesis tree manager for multi-step research investigations. Use when investigating "why X fails" questions that branch into sub-hypotheses, or when you need to track falsifiable claims with linked evidence and timestamps. Subcommands — add,…
Hypothesis-first interpretation of eval results. Use after an eval batch completes, after /post-eval populates analyses, or when preparing paper narrative from result JSONs. Phase 1 requires a prior hypothesis (prevents post-hoc rationalization); Phase 2…
Long-running evaluation campaign with tmux isolation. Use when running any eval batch exceeding ~30 minutes, canonical benchmark runs, or when the user will be away during execution. Pre-flight verification, tmux launch, monitoring, post-flight analysis. NOT…
Zero-context verification that every number, comparison, and scope claim in the paper matches raw result files. Uses a fresh cross-model reviewer with NO prior context to prevent confirmation bias. Use when user says "审查论文数据", "check paper claims", "verify…
Simulate a NeurIPS/SC/ICSE-style peer review with 5 reviewer personas (HPC, ML, Stats, Reproducibility, Devil's Advocate). Use before paper submission, after major methodology changes, or when stress-testing a draft against expected objections. Each reviewer…
Draft LaTeX paper section by section from an outline. Use when user says "写论文", "write paper", "draft LaTeX", "开始写", or wants to generate LaTeX content from a paper plan.
Post-batch analysis pipeline that runs after an eval batch completes. Use after /eval-run or /overnight-eval finishes. Verifies results, runs analysis scripts, refreshes dashboard, writes summary. Does NOT launch new eval runs.
Workflow 4: Submission rebuttal pipeline. Parses external reviews, enforces coverage and grounding, drafts a safe text-only rebuttal under venue limits, and manages follow-up rounds. Use when user says "rebuttal", "reply to reviewers", "ICML rebuttal",…