Coding & Dev Tools

framecoreworks Bundle

Use this skill for provider-neutral narrative structure, story beats, emotional arcs, scene logic, continuity, and multi-shot workflows.

rodrigolmiranda Bundle

Route judgment and mechanical work across Codex subagents and external Claude, Grok and OpenCode sessions. Use for small tasks too when a faster, cheaper worker yields a net gain; preserves explicit ownership, fixed profiles, supervision and independent acceptance.

laabidi-ahmed-ai Skill

A patient, teaching-focused debugging assistant for ML/Python/MLOps errors. Restates the problem, asks clarifying questions only if needed, reasons through root causes ranked by probability (Socratically, guiding before revealing), gives the fix, explains the underlying mechanism, and adds an interview angle. ONLY trigger on explicit invocation by name — e.g. "debug buddy: <error>", "use debug buddy", or "/debug-buddy". Do NOT activate for general questions or errors pasted in normal conversation: if the user hasn't named this skill, do not use it.

laabidi-ahmed-ai Skill

Gives a guided tour of an unfamiliar codebase — a GitHub repo URL or a local code folder — explaining what the project is, how it's structured, which files matter, a "read these first" path, and how to run it, at a student level. ONLY trigger on explicit invocation by name — e.g. "repo onboarding: <url or path>", "use repo onboarding", or "/repo-onboarding <target>". Do NOT activate for general questions or when a repo is merely mentioned: only on explicit naming.

laabidi-ahmed-ai Skill

Runs a live, interactive mock technical interview on an MLOps/ML/DL/software topic — asks ONE question at a time, waits for the user's answer, grades it, probes with follow-ups, and ends with a scorecard of strengths and weak spots. ONLY trigger on explicit invocation by name — e.g. "mock interviewer: <topic>", "use mock interviewer", or "/mock-interviewer <topic>". Do NOT activate for general questions: if the user hasn't named this skill, do not use it.

laabidi-ahmed-ai Skill

Explains a single ML/DL/MLOps concept in a fixed 5-part structure (simple explanation, technical/interview-depth explanation, tiny runnable code example, interview Q&A, common pitfalls). ONLY trigger on explicit invocation by name — e.g. "concept explainer: X", "use concept explainer for X", or "/concept-explainer X". Do NOT activate for general questions on this topic: if the user simply asks about a concept in normal conversation without naming this skill, do not use it.

farzammohammadi Bundle

Manage GitLab Merge Requests - create, review, approve, merge, and full MR lifecycle

farzammohammadi Bundle

Holistic review of completed work to elevate quality from good to excellent. Use after implementation is done to catch what was missed.

farzammohammadi Bundle

Run code, architecture, systems, or proposed changes through a panel of world-class engineering perspectives. Each panelist reads the ACTUAL source files before judging. Use this skill whenever the user asks to: review code quality, get expert opinions, assess architecture decisions, evaluate a refactor plan, critique a design, run something 'through expert eyes', get a 'Linus review', assess quality against the best projects, compare to top OSS standards, or get unbiased multi-perspective feedback. Also trigger for: 'what would X think of this', 'is this good enough', 'how does this compare to the best', 'give me honest feedback', 'tear this apart', 'brutal review', 'no-bullshit assessment'. This is for EVALUATING existing code or proposals — not for writing new code or extracting system structure (use system-layer-extraction for that).

farzammohammadi Bundle

Manage Jira tickets - create, search, update, comment, and transition issues

framecoreworks Bundle

Use this skill for provider-neutral shot language, lens choices, camera movement, lighting, blocking, color, texture, and cinematic direction.

framecoreworks Bundle

Plan provider-neutral tool routing, cost preflight, upload boundaries, approval requirements, fallback limits, and execution risk before any external run.

framecoreworks Bundle

Use this skill to convert direction into beats, scenes, shot cards, timing, transitions, and continuity rules.

viderstar Skill

Design and read A/B tests: sample size and MDE before the start, validity checks on the split, reading the result with a correction for multiple comparisons, and writing the conclusion for a business audience. Use when asked to "compute the MDE", "how much traffic do we need", "analyse this A/B", "look at the test results", "is it significant", "design an experiment", "why did the groups diverge". Not for offline evaluation on historical data, and not for rolling an arm out to production.

viderstar Skill

Query ClickHouse without getting an inflated number or melting a production node: deduplication on ReplacingMergeTree, partition pruning with PREWHERE, column pruning, join order, and how to explore an unfamiliar schema first. Use when asked to "count this from the data", "how many orders/sales/events", "pull that out of ClickHouse", "write a query", "what tables are there", "check this hypothesis against the data", "why is this query hanging", "optimise this query", and before any analysis that starts with "how many do we have". Not for designing tables and migrations, not for A/B test analysis.

viderstar Bundle

Run a Jupyter notebook headless on a remote machine through nbconvert with a named profile of env flags: pre-flight check, launch detached from ssh, notification when it finishes, failure triage and comparison against the previous run. Use when asked to "run this notebook", "run the experiment", "rerun with different parameters", "train on the server", "run the smoke/cheap variant", "watch the run", "tell me when it's done", "why did the notebook fail", "what's the status of the run", and when reproducing a run from an experiment write-up. Not for editing notebooks interactively and not for triggering Airflow DAGs.

johnsmithca-sta Bundle

技能固化与发布工具包。当用户要"把工作流固化为 Skill、起草或完善 SKILL.md、做发布前脱敏与安全预检、打包 SkillHub zip、创建 GitHub tag 保护 ruleset、走双平台（SkillHub/GitHub）发布流程、沉淀可复用方法论、做技能触发词评估或评测循环"时使用。覆盖固化判定、技能目录三件套、SKILL.md 五要素、五层测试闭环（含基线双跑对照）、两轮脱敏审查、发布前 16 项检查清单（含 author/Copyright 归属门禁，按改动类型分级裁剪）、8 脚本自动化（含依赖自检与推送兜底）、触发词评估与评测循环、双平台发布、复盘与自动化反哺。内置脚本零第三方依赖（仅 Python 标准库 + 可选 gh/skillhub CLI），可直接接入 CI 门禁。不适用 / 不用于：技能内部业务逻辑的实现与运行期排障（属发布链路之外的事）、与技能开发无关的一次性脚本、已有成熟流水线且只想跳过检查直接发布的情形。

farzammohammadi Skill

Author a comprehensive, reviewer-ready pull request or GitLab merge request description from the current branch's changes and the working session's context, saved to an isolated temp file. Use this whenever someone is wrapping up a piece of work and needs the writeup — phrasings like "write the PR description", "draft the PR/MR", "create a pull request description", "describe these changes for review", or finishing a branch before opening a merge request. The skill's edge is that it carries the reasoning the diff can't show — why this approach, the root cause behind a bug, the decisions and rejected alternatives, scope and follow-ups, related tickets, and how this PR fits a multi-PR effort.

silkyland Bundle

Set up, tune, and stabilize Linux on an AMD BC-250 board (Cyan Skillfish, the cut-down PS5 APU, PCI 0x13fe / gfx1013). Installs the GPU governor that unpins the 1500 MHz lock, enables CPU C-states and P-states, raises the dynamic VRAM ceiling, and replaces ZRAM with zswap. Use when a user mentions BC250/BC-250, Cyan Skillfish, gfx1013, cyan-skillfish-governor, or reports that such a board stutters, crashes, or sits at a locked 1500 MHz.

s2terminal Skill

git rebaseでブランチを最新化する

s2terminal Skill

review-plannerの計画に従ってコードレビューを実行するスキル

s2terminal Skill

コードレビューする前に計画書を作成するスキルです。レビュー対象のdiff、またはGitの2つのrevisionを入力すると、コードレビューの計画書を作成します。

s2terminal Skill

ChatGPT、Gemini、Claudeに対して、同じ内容を調査させる

s2terminal Skill

Pythonのプログラムやスクリプトを新規作成するときに、通常のPythonプロジェクトかinline script metadata形式かを確認する。単一ファイル形式が選ばれた場合は、PEP 723の依存関係を埋め込み、uvで起動できるPythonスクリプトを作成する。既存のinline metadata scriptの修正にも使う。

framecoreworks Bundle

Use this skill to structure references into canonical sources, aliases, role tags, suppression rules, conflicts, and continuity anchors.

framecoreworks Bundle

Use this skill to coordinate an end-to-end creative video package for reels, Shorts, TikTok, paid social, product films, UGC-style videos, explainers, music-video routes, cutdowns, and local editing, while preserving provider-neutral planning, QA, and delivery gates.

framecoreworks Bundle

Use this skill to review generated or produced outputs against brief, references, prompts, copy locks, expected observables, and acceptance criteria.

framecoreworks Bundle

Use this skill to specify storyboard board, production board, or shot board graphics with panel structure, board copy, labels, and layout constraints.

mrosati84 Skill

Analyze a codebase to understand architecture, dependencies, runnable tasks and runtime information.

mrosati84 Skill

Create elegant, usable, world-class web interfaces; apply when designing or implementing UI/UX, layouts, landing pages, dashboards, forms, or components.

will702 Bundle

Find GitHub repositories through natural semantic search, discover reference implementations, starter templates, and peer projects, or compare similar repositories to help developers build applications. Use whenever a developer asks to find GitHub repos, boilerplates, libraries, open-source alternatives, or architectural references for their tech stack.

organic-san Skill

清理整個專案裡冗長反覆的註解與過時的文件，並找出繞路、多餘的程式碼設計提出改進方案。當使用者要求整理註解、清理專案、檢查註解是否過時或囉唆、更新 README 或 spec、或想知道專案裡有哪些繞圈的設計時，使用這個 skill。註解與文件直接修改，程式碼本身不動、只回報建議。

organic-san Skill

在動手實作之前，先與使用者對焦提案的意圖、代價與替代方案。當使用者提出的內容包含尚未實作的構想、新功能、新模組、架構調整或設計方案，且這輪對話預期最終會導向實作時，務必先使用這個 skill 進行討論，不要直接開始寫程式。純粹的錯誤修正、bug 診斷、現況評估、事實確認與說明性問題不適用。

jtaroreh Bundle

Use for "how does X work", code walkthroughs before changing something, and placement / ownership / layering questions ("where should this live", "which package owns this", "is this the right layer"). Explains subsystem architecture, runtime flow, onboarding mental models. Use why for motivation.

jtaroreh Bundle

Run iterative verification hillclimbs against a concrete test command using Antigravity reactive scheduling. Use for /loop, 'loop until green', or automated hypothesis testing with a verification predicate.

framecoreworks Bundle

Use this skill to translate song context, artist persona, motifs, rhythm, and emotional arc into a music-video direction contract.