Capafy AI - Agent Skills Marketplace
Expert Agents, ready for your task.
Hands-on prompt engineering and LLM evaluation. Built around a diagnose-then-ship loop: forensic transcript critique ▎ that names the exact failure mode, paired with a fixer that produces complete, deployment-ready rewrites — not ▎ vague advice.
Paste anything you don't get. It finds where you're actually stuck — no need to ask the right question — and hands you that "oh, now I get it" moment.
AI prompt architect that turns rough ideas into professional, ready-to-use prompts. It understands your goal, identifies what matters, asks only essential questions, and selects the right prompt architecture for the task. Build, refine, audit, and reuse prompts across leading AI models and tools — with intelligent Commands and Prompt Vault support.
Tell it your commands and gotchas, get a working AGENTS.md or CLAUDE.md back in the first reply. It captures fragile build steps, do-not-touch files, past mistakes and testing gotchas, marks gaps as TODOs and asks only what is missing. Ready for Claude Code, Cursor, Copilot and Codex.
Paste your CLAUDE.md or AGENTS.md and tell it what your agent actually did wrong. It names the rule that failed, quotes the line that should have caught it or says plainly that nothing in your file covers it, and hands you replacement text you can paste back. Built from logged incidents on a production pipeline run with a coding agent, not from general advice. With no incident it still reports what your file itself shows. It makes failures visible; it does not promise they stop.
Most broken skills are not broken. They load, they validate, and they are never invoked - because the description does not match how anyone actually asks. Covers the one-character YAML mistake that makes frontmatter load empty with no error, how to write a description as a match target rather than a summary, and how to prove invocation in two steps instead of hoping.
Turn any website into a guided experience. Analyzes your UI, audits it against ten published onboarding anti-patterns, then generates and Playwright-verifies a spotlight tour. v1.2 adds deep links, mid-tour resume, and a false-bottom scroll cue. Free & MIT.
A kickoff protocol (agent skill) for multi-step tasks. Before writing any code, the agent must: run a spec interview, batch every human-authorization step into one checklist, build a self-verification loop (log sink + verify script + dry-run) *before* the feature itself, and build comparable variants instead of blocking on third-party decisions.
Create a reversible, project-local Codex copy of Claude Code agents, skills, commands, compatible hooks, MCP, and persistent GPT routing—while keeping Claude untouched.
家裡或店裡看到蟑螂，不知道是偶發入侵還是已經繁殖？描述出現位置、大小、顏色、時間、數量與環境線索，蟑螂偵探會協助你判斷可能類型、藏匿位置、入侵來源、風險等級與處理順序。
Catch dangerous Postgres DDL before merge. Paste SQL or ORM migration files and get a GREEN/AMBER/RED verdict, 0-100 score, exact findings, and a merge-gate checklist. Free local Download; no API key or database connection.
Expert in end-to-end AI-assisted software delivery, from design engineering and design systems to shipping: design engineering, design system automation, technical documentation, spec-driven planning, polyglot code quality, systematic QA, release auditing, LLM context engineering, and Astro frontend optimization. Built for agent-agnostic workflows across the full 9-skill delivery lifecycle.
A ready-to-execute cost-reduction kit for infra owners: savings across EC2, RDS, S3, Lambda, network/NAT, and EBS with CLI commands and expected savings, a 44-item monthly review checklist, 15 one-hour quick wins, an executive report template, and a worked case study.
A site architecture and internal linking tool for SaaS, e-commerce, content, docs, local, and subscription sites, designed for new builds, restructures, and SEO audits. Accepts a sitemap, page list, GSC errors, or a simple description. It produces a full structure audit, hierarchy plan, and internal linking recommendations. Orphan pages, misplaced URLs, and weak depth structures are detected, while GSC signals are translated into clear fixes that improve crawlability and help key pages surface.
This agent helps Python teams using the Anthropic or OpenAI SDK safely cut LLM spend without risking regressions. With zero code changes, it maps call sites and runs your test scenarios to rank token waste from missed prompt caching or heavy model tiers. It tests fixes in isolated git worktrees, enforcing a strict "do-no-harm" gate that reverts any patch failing to preserve behavior. It delivers a clear savings audit that converts into verified, behavior-preserving code patches upon opt-in.
Connect Cursor, Claude Code, Hermes & Codex with a shared local SQLite brain. Slash token costs by 92.2% with instant, zero-amnesia cross-agent recall. 100% private, BYOK, lifetime buyout.
In AI manga, your hero is black-haired with square glasses in episode 1 and a brown-haired stranger by episode 4. This keeps the SAME character across all 100 panels — with a feature-lock system, 6 art styles, and the exact fix prompts for all 8 ways AI manga breaks… Free taste first: tell me a concept and I'll give you a full multi-episode story structure before you set up anything. Built by a creator who actually ships AI manga.
Convert object-detection datasets between YOLO, COCO and Pascal VOC in one command, and split any image dataset into train/val/test with stratification and duplicate grouping, so near-identical images never leak between training and test sets. Outputs a ready-to-train data.yaml or COCO JSON. Runs locally, never modifies your source files.