Claude Capability Bridge

Use when running as a third-party or non-default model inside Claude Code, Claude desktop/web, or any Claude-compatible host and the task needs real capabilities - files and git, shell and builds, browsers and web apps, MCP connectors, artifact delivery, GUI control, delegation, or provider/gateway diagnosis. Supplies evidence-first operating procedure - identify the host from live evidence, confirm what is actually exposed and permitted before acting, route to the matching capability card, verify the requested outcome instead of trusting tool success, and report UNKNOWN honestly. Also use whe

37 stars 3 forksAI, LLMs & Agents

Install

skills CLI (Claude Code, Cursor, Codex and more)

npx skills add Abolfazlshahi/claude-capability-bridge-skill --skill claude-capability-bridge

Manual install

git clone --depth 1 https://github.com/Abolfazlshahi/claude-capability-bridge-skill.git /tmp/claude-capability-bridge-skill
mkdir -p ~/.claude/skills
cp -r /tmp/claude-capability-bridge-skill/. ~/.claude/skills/claude-capability-bridge

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

This file is the **kernel**: the small set of rules that stay true in every
host, every session, and every task family. It is written to be stable - it
contains no timestamps, no session state, no live capability claims - so it can
sit unchanged at the front of a conversation.

Everything task-specific lives in `cards/`, `profiles/`, and `references/`, and
is read only when the task needs it.

## What this does and does not do

It reduces one specific gap: a model can *see* tools without knowing their
prerequisites, correct order, or how to prove the result.

It does **not** create tools, grant permissions, connect integrations, fix a
gateway, give the model persistent memory, or enable any provider cache
feature. Text cannot do those things. Claiming otherwise is a failure mode, not
a feature.

## Fast path: do not perform ceremony

If the request can be answered from knowledge and the conversation, with no
tool, no external data, and no side effect:

- **answer directly.** No runtime detection, no probing, no capability report,
  no status vocabulary, no card reading.

If the request needs tools:

- check only the **specific, task-relevant** capability you are unsure about;
- do not re-discover the whole environment, and do not re-verify what this same
  context already established;
- read at most the one card that matches the task family.

Agentic overhead on a simple request is a defect. So is a fresh environment
sweep every turn.

## Kernel rules

Read the full skill on GitHub

Details

Skill name: claude-capability-bridge
Repository: Abolfazlshahi/claude-capability-bridge-skill
License: MIT
Repo updated: 11 September 2026

FAQ

What does the Claude Capability Bridge skill do?

How do I install the Claude Capability Bridge skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add Abolfazlshahi/claude-capability-bridge-skill --skill claude-capability-bridge. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Claude Capability Bridge popular?

It lives in Abolfazlshahi/claude-capability-bridge-skill, which has 37 GitHub stars and 3 forks.

Related skills

Finai Tools Hub

lj22503/awesome-finai-tools-zn

中国金融 AI 工具目录 — 查询/安装/接入 A股/MCP/量化框架。加载此 skill 后，agent 知道每个工具的输入/输出/安装命令，在用户有金融数据需求时自动推荐和安装对应工具。

AI, LLMs & Agents

Booking Best Price

aiwithremy/claude-skills-booking-best-price

Use when the user pastes a holiday-rental or hotel listing link (Airbnb, Stayz, Vrbo, Expedia, Booking.com, HomeToGo, or similar) and wants to identify the actual property, find it listed on other sites, compare prices to book it cheaper, or track down the property manager / owner contact to book direct. Requires the Firecrawl MCP server. This skill verifies Firecrawl is installed and walks the user through setup first if it is not.

AI, LLMs & Agents

Vps Vpn Setup

huasanai/vps-vpn-setup

一键部署 VPS + VPN + Clash 代理环境。适合已购买 VMiss 或 DMIT 服务器、想通过干净 IP 访问 Claude/ChatGPT 等 AI 工具的用户。触发场景： - 用户提供了 VPS 的 IP、密码或 PEM 密钥，要求部署代理环境 - 用户说"帮我装 VPN"、"帮我配置服务器"、"我买了 VMiss/DMIT" - 用户提供了形如 /skill VPS信息: IP: xxx 的格式覆盖能力：SSH 远程执行、v2ray-agent 安装（VLESS Reality）、订阅链接获取、Clash Verge 配置指导、DNS 防泄漏验证。只要用户提供了 VPS 凭据并表达想配代理，ALWAYS invoke this skill。

AI, LLMs & Agents

Agentic Command Center

Forgemind-git/agentic-command-center

An advisory blueprint for planning your own AI command center — one chat or voice interface that runs your whole business by routing to internal systems you own. Guidance and architecture only; it explains the four phases (own your systems, connect your tools, expose them over MCP, unify under one command layer) but never writes code or builds anything. Use when someone wants to understand how to approach internal dashboards, integrations, MCP, and a unified agentic command layer.

AI, LLMs & Agents14

Context Doctor

chrono-meta/forge-harness

Reduces token waste in Claude Code sessions (run the built-in `/context` first; this skill covers what it does not) across two axes — context footprint (auto-generates .claudeignore, guides over-read files and /clear timing) and command output (routes to a command-output proxy/hook such as rtk to trim verbose CLI stdout). In hub environments, regularly audits bloated CLAUDE.md/MEMORY.md/memory/*.md files and proposes compression. Usable standalone without a hub clone.

AI, LLMs & AgentsIn 1 list

Kaoyan Navigator

mcxiaoxiao/kaoyan-navigator-skill

调研中国硕士研究生考试（考研）目标院校与专业的历年招生数据，评估择校爆冷/爆热、缩招、推免挤压、复试线虚低、专业课变更和复试风险，并给出冲稳保建议。适用于用户询问考研择校、院校难度、报录比、复试线、招生计划、统考名额、推免比例、拟录取分数、某校会不会爆、目标分数或需要导入自有 CSV/JSON 数据时。应先搜索公开网络和官方来源，确实找不到关键数据后再向用户索取补充材料；支持 Codex、Claude Code 和其他能读取 Skill Markdown 的 agent。

AI, LLMs & Agents

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.