tcztzy

24 mods across 1 repository, 2 stars between them.

tcztzy/skills

Skill Claude CodeCodex

Translate academic papers, theses, dissertations, abstracts, literature reviews, methods, results, captions, and reviewer responses between Chinese and English with source-faithful completeness, discipline-appropriate terminology, and target-language academic idiom. Use whenever the user asks for 中英互译/英中互译 of journal…

2 1mo ago A 93 tokens

tcztzy/skills

Skill Claude CodeCodex

Implement or debug Agent Client Protocol (ACP) integrations between coding.

2 1mo ago A 0 tokens

benchmark-builder

04

tcztzy/skills

Skill Claude CodeCodex

Build high-quality benchmark items, embedded scoring contracts, review examples, rubrics, evaluators, and release protocols for model and agent evaluation. Use when prompts mention benchmark construction, item writing, rubric design, score-boundary design, checker design, review examples, benchmark health…

2 1mo ago A 71 tokens

bfts-config-prep

05

tcztzy/skills

Skill Claude CodeCodex

Prepare a BFTS run directory from idea JSON by generating idea.md and a configured bftsconfig.yaml. Use when starting a BFTS experiment from structured idea inputs and you want a runnable timestamped run folder under runs/.

2 1mo ago A 54 tokens

code-simplifier

06

tcztzy/skills

Skill Claude CodeCodex

Safely simplify code with multi-metric quality gates. Use when prompts mention to refactor/compress/reduce size or complexity without functional change; validate with formatter-aligned metrics (lines/bytes + tokens/complexity when available) plus lint/tests/build and performance regression checks when relevant.

2 1mo ago A 63 tokens

tcztzy/skills

Skill Claude CodeCodex

Run the standalone BFTS experiment pipeline to execute multi-agent tree-search experiments from a prepared bftsconfig.yaml. Use after bfts-config-prep to produce logs, workspaces, and experiment artifacts.

2 1mo ago A 47 tokens

tcztzy/skills

Skill Claude CodeCodex

Summarize an experiment run directory into a draft report (Markdown + JSON) by scanning logs, summary JSONs, figures, and artifacts without executing code. Use when you have a run folder (AI-Scientist, ML experiments, benchmarks) and need a grounded, source-linked summary for writeups or debugging.

2 1mo ago A 70 tokens

hpc-batch

09

tcztzy/skills

Skill Claude CodeCodex

Troubleshoot Slurm/LSF batch jobs, arrays, and stuck-job issues. Use when prompts mention sbatch/squeue/sacct/scontrol or bsub/bjobs/bhist/bacct, job arrays, N inputs producing fewer than N outputs, missing per-task logs/exit codes, stuck pending/running jobs, overload/queue saturation, safe retry/resubmit decisions…

2 1mo ago A 87 tokens

latex-to-word

10

tcztzy/skills

Skill Claude CodeCodex

Convert LaTeX (.tex) files to Word (.docx) using an automated preprocessing pipeline that resolves common pandoc incompatibilities such as custom commands, tabularx, and resizebox, minimizing manual cleanup afterward.

2 1mo ago A 48 tokens

modern-python-style

11

tcztzy/skills

Skill Claude CodeCodex

Enforce modern Python 3.10+ syntax, static typing discipline, and anti-ceremony code style with deterministic checks. Use when prompts mention banning typing.List or Optional, banning from future import annotations, treating Python like a statically typed language, enabling multiple type checkers, rejecting pointless…

2 1mo ago A 83 tokens

nix

12

tcztzy/skills

Skill Claude CodeCodex

Write, review, and debug Nix expressions, flakes, NixOS/Home Manager/nix-darwin modules, overlays, and derivations. Use for .nix files; Nix evaluation or build failures; package, app, devShell, and system outputs; or nix build/run/develop/flake/config commands.

2 1mo ago A 69 tokens

paper-reviewer

13

tcztzy/skills

Skill Claude CodeCodex

Review academic papers (PDF/LaTeX/Markdown) with a ScholarEval-inspired dimension pass and produce both structured peer-review JSON (NeurIPS-style fields) and a concise human-readable review with prioritized, actionable feedback. Use for comprehensive, targeted, or comparative paper evaluation while keeping output…

2 1mo ago A 64 tokens

paper-visualizer

14

tcztzy/skills

Skill Claude CodeCodex

Route publication-grade visualization tasks by chart family and execute manuscript-facing plot or diagram workflows with packaged agents, shared generators, style guides, and inspectable run artifacts. Use when prompts mention paper figures, chart choice or critique, plotting backends such as matplotlib, seaborn…

2 1mo ago A 108 tokens

paper-writer

15

tcztzy/skills

Skill Claude CodeCodex

Single entry point for drafting or revising scientific papers, generating profile-based LaTeX scaffolds, harvesting draft-local citations into BibTeX, and auditing figure-caption-figref alignment from paper PDFs. Use when writing or editing research papers, workshop submissions, technical reports, negative-results…

2 1mo ago A 151 tokens

project-simplify

16

tcztzy/skills

Skill Claude CodeCodex

Reduce repository complexity by de-vendoring large third-party sources and replacing them with pinned release artifacts, checksum verification, and minimal subset extraction. Use when Codex needs to shrink Git history, migrate external fixtures (especially LLVM/Clang tests), design artifact lock manifests, or build…

2 1mo ago A 68 tokens

raster-to-vector

17

tcztzy/skills

Skill Claude CodeCodex

Convert raster architecture diagrams, screenshots, flowcharts, UI diagrams, scanned figures, charts, and rendered PDF pages into presentation-ready editable SVG. Use PP-OCR text geometry, local segmentation, CV primitives, optional image understanding, agent-provided icon replacements, and per-element fallback…

2 1mo ago A 75 tokens

tcztzy/skills

Skill Claude CodeCodex

Generate and refine research ideas and run a lightweight novelty sanity-check using Semantic Scholar. Use when drafting proposal-style ideas from a topic brief, validating idea JSON outputs, or checking whether an idea seems close to existing work (without over-claiming novelty).

2 1mo ago A 59 tokens

tcztzy/skills

Skill Claude CodeCodex

Decide whether a research idea, experiment thread, or paper draft is worth pursuing and how to sharpen it into one high-impact claim. Use when judging if a project should go, kill, pivot, or scope down; focusing experiments around a single thesis; or repositioning a paper's title, introduction, figures, and conclusion…

2 1mo ago A 73 tokens

skill-manager

20

tcztzy/skills

Skill Claude CodeCodex

Manage and evolve Codex skills: create, update, and audit SKILL.md files together with their scripts, references, and assets; inventory installed skills; validate and convert Claude Code skills; and check shared runtime readiness. Use when prompts mention create skill, audit skill, update SKILL.md, validate skill…

2 1mo ago C 80 tokens

token-cost-tracker

21

tcztzy/skills

Skill Claude CodeCodex

Summarize token usage and cost records from JSON or JSONL logs.

2 1mo ago A 20 tokens

typst

22

tcztzy/skills

Skill Claude CodeCodex

Draft, convert, debug, and validate Typst documents. Use when working with .typ or .typst files, migrating LaTeX snippets to Typst, fixing Typst syntax, checking rendered output, or authoring reports, slides, posters, and templates where typst-mcp can provide docs lookup, conversion, syntax checks, and snippet…

2 1mo ago A 88 tokens

wsl-fs

23

tcztzy/skills

Skill Claude CodeCodex

Use when the task needs to work with projects stored in the WSL Linux filesystem, including path translation, entering a distro, finding repos under /home, and running build or git commands in WSL instead of on Windows.

2 1mo ago A 50 tokens

zotero

24

tcztzy/skills

Skill Claude CodeCodex

Capture (Zotero Connector), manage, search, and export Zotero libraries and collection-scoped citations using Zotero desktop + Better BibTeX, the Zotero Web API, a local Zotero data directory, or user-provided export files. Use when tasks mention Zotero, Zotero Connector, Better BibTeX, collections, keep-updated…

2 1mo ago A 124 tokens