Recently updated open-source projects

expert-debate turns an idea into two deliberately conflicting professional proposals instead of a balanced pros-and-cons list. Claude writes from one expert role, then produces a copyable prompt for an independent Model B session with the opposing role and no shared context. You compare the arguments and record your reasoning manually, so the workflow needs no API keys or automated model integration.

blind-review prepares a prompt for a fresh Model B session that evaluates a document without knowing the author’s intent or prior collaboration context. Comparing that independent reading with what you meant can reveal missing assumptions and “knowledge curse” gaps before a PRD or specification reaches stakeholders. An optional devil’s-advocate pass adds pressure when reviewers agree too quickly, and the workflow uses manual copy-paste rather than an API.

doc-generator converts a finalized product idea into development-ready documentation, including a PRD, technical specification, CLAUDE.md, and implementation guide. Its PRD flow includes an AI-simulated user-validation step, and the command supports a full four-document suite, PRD-only output, a unified PRD plus blueprint, or an interactive from-scratch mode. This lets teams and solo builders choose the amount of structure they need.

Complete an eight-question conversational diagnostic to identify where your contextual judgment, trust, lived pattern recognition, ambiguity handling, or human presence stands out. The skill produces a specific profile, practical visibility changes, audience-ready positioning language, and a professional post lead. You can keep logging experiences and periodically refresh the profile as your evidence and context evolve.

Search an indexed set of 80 UI-relevant Nano Banana prompts, then adapt a selected pattern for Gemini image generation, Google Stitch screens and design systems, or coded React, Tailwind, and SwiftUI work. The skill extracts composition, palette, and typography details instead of blindly copying image language. Attribution must be preserved, and the included English collection is only a curated subset of the upstream library.

Give it a rough idea, broken instruction, or underperforming prompt. The skill classifies the request, diagnoses missing context or structure, selects relevant prompting techniques, and returns a ready-to-use rewrite with explicit output requirements and guardrails where needed. It also explains what changed and which techniques were applied or intentionally skipped, so the result doubles as a learning aid.

Provide a product URL, feature list, research notes, or roadmap and receive an interactive HTML map connecting customer outcomes, needs, business goals, current solutions, confidence, and competitive context. Teams can use it in working sessions to distinguish essential value from differentiators, edit or rearrange nodes, track changes, and save a self-contained snapshot for later discussion.

Activate a Spanish brevity mode that removes greetings, hedging, redundant pronouns, copulas, and long connective phrases while preserving technical terms and actionable fixes. It is designed to reduce explanatory output substantially without changing the underlying solution. Code, quoted errors, commit messages, pull requests, and formal technical documentation remain in their normal form.

Atomic Emails guides a SaaS campaign through a structured five-step method built around one email per idea. It gathers the audience, campaign goal, success steps, obstacles, positioning, and existing resources, then turns that material into a storyboard and finished copy. The workflow fits onboarding, trial expiry, customer success, lead nurture, and re-engagement campaigns, and can start from an existing brief or storyboard.

Describe a page, API, or form and the skill will analyze its fields, validation rules, and dependencies before defining valid and invalid equivalence classes. It applies the CASCADE pattern to plan coverage with a compact set of test cases, then generates a formatted Excel workbook with formulas and coverage metrics. The output is compatible with Nanook's ImporterXlsx format.

deliverable puts an idea through three adversarial perspectives—a cynical customer, a brutal product manager, and a future maintainer—then writes a one-page verdict with fatal assumptions, a narrower wedge, and evidence to gather next. Each project keeps its own JSONL history, so related ideas can be challenged with previous lessons. Ideas that survive can continue into a charter, BRD, SRS, and stakeholder-interview workflow.

Claude LTM stores project decisions in `.memory/DECISIONS.md`, with separate detail files for complex choices. Claude Code is instructed to read this memory at session start, update it when significant decisions are made, and compact the session before finishing. The plugin provides init, save, and compact commands without external services or hooks.

Convert CSV, JSON, NDJSON, YAML, SQL, Markdown, TOML, and related formats through Claude Code or the bundled Python scripts. It can flatten nested records, stream large files, generate SQL statements, and infer column types or schemas. The core tools use Python 3.8 standard-library modules, with PyYAML needed only for YAML support.

Use /ick when Claude’s output is technically correct but stylistically wrong, such as over-commented code, corporate filler, needless abstraction, hedging, or sycophantic openings. The skill identifies the specific problem, rewrites the offending output, and saves the preference to memory. With no pasted snippet, it inspects the most recent response and fixes the most likely source of discomfort.

Growth Coach helps technical founders diagnose problems across product-market fit, acquisition, pricing, sales, fundraising, and founder wellbeing. It selects an appropriate framework, proposes three to five time-boxed actions, and includes a usable deliverable such as a cold email, LinkedIn post, or pricing table; the first run registers a free API key with 50 queries.

Provide a Bilibili, YouTube, podcast, WeChat, Zhihu, or long-form article link—or a local audio or video file—and turn the material into a formatted ebook. It is designed for knowledge workers organizing saved content, students and researchers processing lectures, and creators repurposing long-form material into something easier to read and reuse.

Provide the target company, role, job description, and your resume to create a focused software-engineering interview kit. The skill produces five linked Markdown guides covering a day-by-day plan, prioritized DSA practice, resume-based behavioral prompts, questions for each interview round, and company stack topics. It also includes short-on-time checklists so candidates can use the material immediately instead of reading a generic handbook.

Give the skill your LinkedIn profile PDF, photo, and examples you admire, and it will score the profile section by section, rewrite weak copy, and create a niche-specific editorial calendar. It can also prepare posts with PNG images or carousels and adapt future batches from engagement data you provide. Publishing remains manual by default; image generation and scheduled posting require optional Gemini or n8n setup.