opencode-skills-collection 3.0.13 → 3.0.15

package/bundled-skills/.antigravity-install-manifest.json

@@ -1,6 +1,6 @@
 {
   "schemaVersion": 1,
-  "updatedAt": "2026-05-13T01:56:16.789Z",
+  "updatedAt": "2026-05-15T01:57:26.024Z",
   "entries": [
     "00-andruia-consultant",
     "007",
@@ -855,6 +855,7 @@
     "maxia",
     "mcp-builder",
     "mcp-builder-ms",
+    "mcp-tool-developer",
     "memory-forensics",
     "memory-safety-patterns",
     "memory-systems",
@@ -915,6 +916,7 @@
     "network-engineer",
     "networkx",
     "new-rails-project",
+    "news-sentiment-engine",
     "nextjs-app-router-patterns",
     "nextjs-best-practices",
     "nextjs-supabase-auth",
@@ -968,6 +970,7 @@
     "one-drive-automation",
     "openapi-spec-generation",
     "openclaw-github-repo-commander",
+    "options-flow-analyzer",
     "oral-health-analyzer",
     "orchestrate-batch-refactor",
     "os-scripting",
@@ -1326,6 +1329,7 @@
     "tiktok-automation",
     "tmux",
     "todoist-automation",
+    "tokenwise",
     "tool-design",
     "tool-use-guardian",
     "top-web-vulnerabilities",
@@ -1409,6 +1413,7 @@
     "web3-testing",
     "webapp-testing",
     "webflow-automation",
+    "wechat-official-account-strategist",
     "weightloss-analyzer",
     "wellally-tech",
     "whatsapp-automation",
@@ -1438,6 +1443,7 @@
     "x-article-publisher-skill",
     "x-twitter-scraper",
     "x402-express-wrapper",
+    "xiaohongshu-content-strategist",
     "xlsx-official",
     "xss-html-injection",
     "xvary-stock-research",

package/bundled-skills/docs/integrations/jetski-cortex.md

@@ -1,9 +1,9 @@
 ---
 title: Jetski/Cortex + Gemini Integration Guide
-description: "Use antigravity-awesome-skills with Jetski/Cortex without hitting context-window overflow with 1,453+ skills."
+description: "Use antigravity-awesome-skills with Jetski/Cortex without hitting context-window overflow with 1,459+ skills."
 ---
 
-# Jetski/Cortex + Gemini: safe integration with 1,453+ skills
+# Jetski/Cortex + Gemini: safe integration with 1,459+ skills
 
 This guide shows how to integrate the `antigravity-awesome-skills` repository with an agent based on **Jetski/Cortex + Gemini** (or similar frameworks) **without exceeding the model context window**.
 
@@ -23,7 +23,7 @@ Never do:
 - concatenate all `SKILL.md` content into a single system prompt;
 - re-inject the entire library for **every** request.
 
-With 1,453+ skills, this approach fills the context window before user messages are even added, causing truncation.
+With 1,459+ skills, this approach fills the context window before user messages are even added, causing truncation.
 
 ---
 

package/bundled-skills/docs/integrations/jetski-gemini-loader/README.md

@@ -20,7 +20,7 @@ This example shows one way to integrate **antigravity-awesome-skills** with a Je
 - How to enforce a **maximum number of skills per turn** via `maxSkillsPerTurn`.
 - How to choose whether to **truncate or error** when too many skills are requested via `overflowBehavior`.
 
-This pattern avoids context overflow when you have 1,453+ skills installed.
+This pattern avoids context overflow when you have 1,459+ skills installed.
 
 ---
 

package/bundled-skills/docs/maintainers/repo-growth-seo.md

@@ -6,7 +6,7 @@ This document keeps the repository's GitHub-facing discovery copy aligned with t
 
 Preferred positioning:
 
-> Installable GitHub library of 1,453+ agentic skills for Claude Code, Cursor, Codex CLI, Gemini CLI, Antigravity, and other AI coding assistants.
+> Installable GitHub library of 1,459+ agentic skills for Claude Code, Cursor, Codex CLI, Gemini CLI, Antigravity, and other AI coding assistants.
 
 Key framing:
 
@@ -20,7 +20,7 @@ Key framing:
 
 Preferred description:
 
-> Installable GitHub library of 1,453+ agentic skills for Claude Code, Cursor, Codex CLI, Gemini CLI, Antigravity, and more. Includes installer CLI, bundles, workflows, and official/community skill collections.
+> Installable GitHub library of 1,459+ agentic skills for Claude Code, Cursor, Codex CLI, Gemini CLI, Antigravity, and more. Includes installer CLI, bundles, workflows, and official/community skill collections.
 
 Preferred homepage:
 
@@ -28,7 +28,7 @@ Preferred homepage:
 
 Preferred social preview:
 
-- use a clean preview image that says `1,453+ Agentic Skills`;
+- use a clean preview image that says `1,459+ Agentic Skills`;
 - mention Claude Code, Cursor, Codex CLI, and Gemini CLI;
 - avoid dense text and tiny logos that disappear in social cards.
 

package/bundled-skills/docs/maintainers/skills-update-guide.md

@@ -69,7 +69,7 @@ For manual updates, you need:
 The update process refreshes:
 - Skills index (`skills_index.json`)
 - Web app skills data (`apps\web-app\public\skills.json`)
-- All 1,453+ skills from the skills directory
+- All 1,459+ skills from the skills directory
 
 ## When to Update
 

package/bundled-skills/docs/users/bundles.md

@@ -673,4 +673,4 @@ Found a skill that should be in a bundle? Or want to create a new bundle? [Open
 
 ---
 
-_Last updated: March 2026 | Total Skills: 1,453+ | Total Bundles: 37_
+_Last updated: March 2026 | Total Skills: 1,459+ | Total Bundles: 37_

package/bundled-skills/docs/users/claude-code-skills.md

@@ -12,7 +12,7 @@ Install the library into Claude Code, then invoke focused skills directly in the
 
 ## Why use this repo for Claude Code
 
-- It includes 1,453+ skills instead of a narrow single-domain starter pack.
+- It includes 1,459+ skills instead of a narrow single-domain starter pack.
 - It supports the standard `.claude/skills/` path and the Claude Code plugin marketplace flow.
 - It also ships generated bundle plugins so teams can install focused packs like `Essentials` or `Security Developer` from the marketplace metadata.
 - It includes onboarding docs, bundles, and workflows so new users do not need to guess where to begin.

package/bundled-skills/docs/users/gemini-cli-skills.md

@@ -12,7 +12,7 @@ Install into the Gemini skills path, then ask Gemini to apply one skill at a tim
 
 - It installs directly into the expected Gemini skills path.
 - It includes both core software engineering skills and deeper agent/LLM-oriented skills.
-- It helps new users get started with bundles and workflows rather than forcing a cold start from 1,453+ files.
+- It helps new users get started with bundles and workflows rather than forcing a cold start from 1,459+ files.
 - It is useful whether you want a broad internal skill library or a single repo to test many workflows quickly.
 
 ## Install Gemini CLI Skills

package/bundled-skills/docs/users/getting-started.md

@@ -1,4 +1,4 @@
-# Getting Started with Antigravity Awesome Skills (V11.1.0)
+# Getting Started with Antigravity Awesome Skills (V11.2.0)
 
 **New here? This guide will help you supercharge your AI Agent in 5 minutes.**
 

package/bundled-skills/docs/users/kiro-integration.md

@@ -18,7 +18,7 @@ Kiro is AWS's agentic AI IDE that combines:
 
 Kiro's agentic capabilities are enhanced by skills that provide:
 
-- **Domain expertise** across 1,453+ specialized areas
+- **Domain expertise** across 1,459+ specialized areas
 - **Best practices** from Anthropic, OpenAI, Google, Microsoft, and AWS
 - **Workflow automation** for common development tasks
 - **AWS-specific patterns** for serverless, infrastructure, and cloud architecture

package/bundled-skills/docs/users/usage.md

@@ -14,7 +14,7 @@ If you came in through a **Claude Code** or **Codex** plugin instead of a full l
 
 When you ran `npx antigravity-awesome-skills` or cloned the repository, you:
 
-✅ **Downloaded 1,453+ skill files** to your computer (default: `~/.gemini/antigravity/skills/`; or a custom path like `~/.agent/skills/` if you used `--path`)
+✅ **Downloaded 1,459+ skill files** to your computer (default: `~/.gemini/antigravity/skills/`; or a custom path like `~/.agent/skills/` if you used `--path`)
 ✅ **Made them available** to your AI assistant  
 ❌ **Did NOT enable them all automatically** (they're just sitting there, waiting)
 
@@ -34,7 +34,7 @@ Bundles are **curated groups** of skills organized by role. They help you decide
 
 **Analogy:**
 
-- You installed a toolbox with 1,453+ tools (✅ done)
+- You installed a toolbox with 1,459+ tools (✅ done)
 - Bundles are like **labeled organizer trays** saying: "If you're a carpenter, start with these 10 tools"
 - You can either **pick skills from the tray** or install that tray as a focused marketplace bundle plugin
 
@@ -212,7 +212,7 @@ Let's actually use a skill right now. Follow these steps:
 
 ## Step 5: Picking Your First Skills (Practical Advice)
 
-Don't try to use all 1,453+ skills at once. Here's a sensible approach:
+Don't try to use all 1,459+ skills at once. Here's a sensible approach:
 
 If you want a tool-specific starting point before choosing skills, use:
 
@@ -343,7 +343,7 @@ Usually no, but if your AI doesn't recognize a skill:
 
 ### "Can I load all skills into the model at once?"
 
-No. Even though you have 1,453+ skills installed locally, you should **not** concatenate every `SKILL.md` into a single system prompt or context block.
+No. Even though you have 1,459+ skills installed locally, you should **not** concatenate every `SKILL.md` into a single system prompt or context block.
 
 The intended pattern is:
 

package/bundled-skills/docs/users/visual-guide.md

@@ -34,7 +34,7 @@ antigravity-awesome-skills/
 ├── 📄 CONTRIBUTING.md                  ← Contributor workflow
 ├── 📄 CATALOG.md                       ← Full generated catalog
 │
-├── 📁 skills/                          ← 1,453+ skills live here
+├── 📁 skills/                          ← 1,459+ skills live here
 │   │
 │   ├── 📁 brainstorming/
 │   │   └── 📄 SKILL.md                 ← Skill definition
@@ -47,7 +47,7 @@ antigravity-awesome-skills/
 │   │   └── 📁 2d-games/
 │   │       └── 📄 SKILL.md             ← Nested skills also supported
 │   │
-│   └── ... (1,453+ total)
+│   └── ... (1,459+ total)
 │
 ├── 📁 apps/
 │   └── 📁 web-app/                     ← Interactive browser
@@ -100,7 +100,7 @@ antigravity-awesome-skills/
 
 ```
                     ┌─────────────────────────┐
-                    │  1,453+ SKILLS          │
+                    │  1,459+ SKILLS          │
                     └────────────┬────────────┘
                                  │
         ┌────────────────────────┼────────────────────────┐
@@ -201,7 +201,7 @@ If you want a workspace-style manual install instead, cloning into `.agent/skill
 │   ├── 📁 brainstorming/                 │
 │   ├── 📁 stripe-integration/            │
 │   ├── 📁 react-best-practices/          │
-│   └── ... (1,453+ total)                │
+│   └── ... (1,459+ total)                │
 └─────────────────────────────────────────┘
 ```
 

package/bundled-skills/mcp-tool-developer/SKILL.md

@@ -0,0 +1,132 @@
+---
+name: mcp-tool-developer
+description: "Build Model Context Protocol (MCP) servers and tools from scratch. Full-stack MCP development with TypeScript/Python, testing, deployment, and registry publishing."
+category: developer-tools
+risk: safe
+source: community
+source_repo: demo112/yunqu-ai-skills
+source_type: community
+date_added: "2026-05-13"
+author: yundu-ai
+tags: [mcp, ai-agent, tool-development, typescript, python, llm, model-context-protocol]
+tools: [claude, cursor, gemini]
+---
+
+# MCP Tool Developer
+
+## Overview
+
+Expert at building Model Context Protocol (MCP) servers that give AI agents new capabilities. Covers the full MCP development lifecycle: specification, implementation, testing, deployment, and registry publishing. Supports both TypeScript and Python with production-ready patterns.
+
+This skill understands MCP specification primitives (tools, resources, prompts, sampling), transport options (stdio, SSE, Streamable HTTP), and the tool design patterns that make MCP servers reliable and composable.
+
+## When to Use This Skill
+
+- Use when building a new MCP server from scratch
+- Use when wrapping an existing API as an MCP tool
+- Use when debugging MCP server issues
+- Use when designing the tool schema for an MCP server
+- Use when publishing an MCP server to a registry
+
+## How It Works
+
+### Step 1: Define the MCP Server Scope
+
+Identify what capabilities the server should expose:
+- **Tools** - Functions the LLM can call (primary use case)
+- **Resources** - Data the LLM can read (files, APIs, databases)
+- **Prompts** - Reusable prompt templates
+
+Choose the transport:
+- **stdio** - For local CLI tools (Claude Code, Cursor)
+- **SSE (Server-Sent Events)** - For remote/hosted tools
+- **Streamable HTTP** - New in MCP spec for modern deployments
+
+### Step 2: Design the Tool Schema
+
+Define input/output schemas before writing implementation:
+
+```typescript
+{
+  name: "tool_name",
+  description: "What this tool does (visible to the LLM)",
+  inputSchema: {
+    type: "object",
+    properties: { ... },
+    required: [ ... ]
+  }
+}
+```
+
+### Step 3: Implement the Server
+
+Create the server with proper error handling, validation, and logging. Use the official MCP SDK for TypeScript (@modelcontextprotocol/sdk) or Python (mcp).
+
+### Step 4: Test and Deploy
+
+Test with the MCP Inspector, validate tool schemas, handle edge cases, then deploy locally or remotely.
+
+## Examples
+
+### Example 1: TypeScript MCP Server
+
+```typescript
+import { McpServer } from "@modelcontextprotocol/sdk/server/mcp.js";
+import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js";
+import { z } from "zod";
+
+const server = new McpServer({ name: "my-tools", version: "1.0.0" });
+
+server.tool("greet", "Greet someone by name",
+  { name: z.string().describe("Person's name") },
+  async ({ name }) => ({ content: [{ type: "text", text: `Hello, ${name}!` }] })
+);
+
+const transport = new StdioServerTransport();
+await server.connect(transport);
+```
+
+### Example 2: API Wrapper Pattern
+
+Wrap an external API as an MCP tool with auth, rate limiting, and error handling:
+- Map API endpoints to tools
+- Handle auth via environment variables
+- Transform API responses to LLM-friendly format
+- Add retry logic with exponential backoff
+
+## Best Practices
+
+- Build small, focused tools that can be chained rather than monolithic tools
+- Return structured errors, not crashes - tools should fail gracefully
+- Define schemas before implementation
+- Include descriptions that help the LLM understand when and how to use each tool
+- Validate all inputs against the schema
+- Add rate limiting for external API calls
+- Use environment variables for secrets, never hardcode credentials
+
+## Limitations
+
+- This skill provides guidance and code generation; actual runtime testing requires a development environment
+- MCP specification is evolving; always check the latest spec version
+- Security review is essential before deploying tools that handle sensitive data
+
+## Security and Safety Notes
+
+- Never hardcode API keys or credentials in tool implementations
+- Use environment variables or secret managers for all authentication
+- Validate and sanitize all inputs to prevent injection attacks
+- Rate limit external API calls to prevent abuse
+- Review tool permissions carefully - tools can access files, networks, and execute code
+
+## Common Pitfalls
+
+- **Problem:** LLM calls tools with wrong parameters
+  **Solution:** Improve tool descriptions and add examples in the description field. The LLM reads descriptions to decide how to call tools.
+
+- **Problem:** Tool times out on large inputs
+  **Solution:** Add input size validation and pagination. Stream large responses instead of buffering.
+
+## Related Skills
+
+- `api-integration-architect` - For API design patterns used in MCP tools
+- `security-audit-code-reviewer` - For reviewing MCP server code security

package/bundled-skills/news-sentiment-engine/SKILL.md

@@ -0,0 +1,104 @@
+---
+name: news-sentiment-engine
+description: Multi-source RSS news aggregation with Claude-powered sentiment analysis and structured briefing output
+category: research
+risk: safe
+source: community
+source_repo: tellmefrankie/news-engine
+source_type: community
+date_added: "2026-05-13"
+author: tellmefrankie
+tags: [news, rss, sentiment-analysis, briefing, research]
+tools: [claude, websearch]
+---
+# News Sentiment Engine (Free)
+
+Collect and analyze AI/tech news from multiple sources with Claude-powered sentiment analysis. Open source lite version.
+
+## When to Use
+
+- Use when preparing a concise AI or technology news briefing from multiple RSS sources.
+- Use when you need ranked article summaries with sentiment, tags, and impact scoring.
+- Use when monitoring industry changes across product launches, policy moves, and infrastructure shifts.
+- Use when deduplicating overlapping coverage before writing a daily or weekly briefing.
+
+## What it does
+
+- Collects news from 4+ RSS feeds (TechCrunch, The Verge, Ars Technica, Hacker News)
+- Deduplicates articles across sources
+- Ranks by importance (industry impact, technology trends, policy changes)
+- Generates structured briefing with sentiment tags
+- Outputs formatted briefing card
+
+## Usage
+
+```
+Collect latest AI/tech news from RSS feeds.
+Rank top 5 by importance to the tech industry.
+For each: summary (2-3 sentences), sentiment (positive/negative/neutral),
+impact score (1-5), industry tags, one-sentence commentary.
+Output as a structured briefing card.
+```
+
+## Example Output
+
+```
+AI/Tech News Briefing — 2026-05-13
+
+1. OpenAI announces GPT-5 with 2M context window
+   Source: TechCrunch | Impact: 5/5
+   Tags: #AI #LLM #OpenAI
+   Sentiment: Positive
+
+   Summary: OpenAI unveiled GPT-5 with a 2M token context window and
+   improved reasoning. Enterprise pricing starts at $0.03/1k tokens.
+
+   Commentary: Direct competitive pressure on Anthropic Claude 3.5.
+   Enterprise deals may shift in H2 2026.
+
+2. EU AI Act enforcement begins for high-risk systems
+   Source: The Verge | Impact: 4/5
+   Tags: #Regulation #EU #Compliance
+   Sentiment: Neutral
+```
+
+## Output Format
+
+For each article:
+- Title + source + publish date
+- Summary (2-3 sentences)
+- Industry tags: [AI, Semiconductor, Cloud, etc.]
+- Sentiment: Positive/Negative/Neutral
+- Impact score: 1-5
+- Commentary: 1-sentence industry perspective
+
+## Setup
+
+```bash
+git clone https://github.com/tellmefrankie/news-engine
+cd news-engine
+pnpm install
+cp .env.example .env
+# Requires: ANTHROPIC_API_KEY
+pnpm dev -- --collect-only
+```
+
+No paid APIs required for free tier. Anthropic API key only.
+
+## Limitations
+
+- RSS feeds can lag, disappear, throttle, or duplicate syndicated coverage.
+- Sentiment and impact scores are briefing aids, not authoritative market or policy analysis.
+- The example setup runs third-party code; review the repository and environment variables before use.
+- Outputs should be cross-checked against original article sources before publication or investment use.
+
+## Pro Version
+
+Free tier covers news collection and basic analysis.
+
+**Full bundle — $29 one-time**: Investment-grade analysis (portfolio impact scoring, options flow correlation, earnings catalyst detection), Telegram auto-delivery.
+→ https://jaehyunpark.gumroad.com/l/tcyahy
+
+## Author
+
+Core module from a production news analysis engine processing 50+ articles daily since 2026.

package/bundled-skills/options-flow-analyzer/SKILL.md

@@ -0,0 +1,96 @@
+---
+name: options-flow-analyzer
+description: Real vs lottery call separation for options P/C ratio analysis — prevents signal inversion from deep OTM noise
+category: finance
+risk: safe
+source: community
+source_type: community
+date_added: "2026-05-13"
+author: tellmefrankie
+tags: [options, sentiment-analysis, trading, polygon, market-analysis]
+tools: [websearch]
+---
+# Options Flow Analyzer
+
+Analyze options chain data with real vs lottery call separation — the key insight that prevents P/C ratio misinterpretation. Uses Polygon.io API.
+
+## When to Use
+
+- Use when raw put/call ratios appear bullish or bearish but may be distorted by cheap deep OTM contracts.
+- Use when comparing options flow across watchlists, holdings, sectors, or event-driven names.
+- Use when you need to separate institutional hedging from speculative lottery-ticket activity.
+- Use when tracking options anomalies against a recent baseline.
+
+## What it does
+
+Standard P/C ratio analysis is misleading. A P/C of 0.35 looks "extremely bullish" but may be 84% lottery calls ($0.01-$0.09 OTM options).
+
+This skill separates:
+- **Real calls**: Strike price within 5% of stock price, meaningful premium
+- **Lottery calls**: Deep OTM, cheap premium, speculative bets
+- **Real puts**: Actual hedging activity
+- **Lottery puts**: Cheap downside bets
+
+## Analysis Output
+
+For each ticker:
+- Real P/C ratio (excludes lottery noise)
+- Lottery percentage (what % of volume is speculation)
+- Per-expiry breakdown (weekly vs monthly vs LEAPS)
+- Anomaly detection: P/C shifts >0.3, Call OI surges >30%, IV spikes >20%
+- Sentiment classification: Bullish/Bearish/Neutral with confidence
+
+## Example Output
+
+```
+Options Flow Summary — 2026-05-13
+
+HOLDINGS:
+CEG  $299.69 | Raw P/C: 1.06 | Lottery: 61% | Adj P/C: 2.72  BEARISH (was neutral raw)
+IREN $55.15  | Raw P/C: 0.83 | Lottery: 34% | Adj P/C: 0.55  BULLISH
+KTOS $56.99  | Raw P/C: 0.53 | Lottery: 28% | Adj P/C: 0.38  EXTREME BULLISH
+RXRX $3.26   | Raw P/C: 0.38 | Lottery: 84% | Adj P/C: 2.37  BEARISH (was extreme bullish raw)
+
+SECTORS:
+XLI  | Raw P/C: 5.32 | Lottery:  8% | Adj P/C: 4.89  INSTITUTIONAL HEDGE
+
+ANOMALIES:
+XLI: P/C 5.32 vs 30-day baseline 0.87 — 4.5 std deviations above normal
+RXRX: 84% lottery calls — raw P/C signal completely inverted after filtering
+```
+
+## Configuration
+
+```
+Analyze options flow for my watchlist:
+Holdings: CEG, IREN, KTOS, RXRX, TEM
+Sectors: SPY, QQQ, XLI, XLK
+Separate real vs lottery calls (threshold: premium < $0.10, delta < 0.05).
+Flag anomalies vs 30-day baseline.
+```
+
+## Requirements
+
+- Polygon.io API key (free tier covers basic data; paid tier for full chain)
+- WebSearch for cross-verification
+
+## Limitations
+
+- Options data can be delayed, incomplete, or unavailable depending on the Polygon.io plan.
+- Heuristics such as premium and delta thresholds need adjustment for ticker price, volatility, and expiry.
+- Sentiment classifications are analytical signals, not financial advice or trade recommendations.
+- Always cross-check unusual flow against price action, news catalysts, liquidity, and risk controls.
+
+## Key Discovery
+
+This real/lottery separation was discovered during live portfolio management when RXRX showed P/C 0.35 (looks extremely bullish) but was actually 84% lottery calls at $0.01-$0.09. The "bullish signal" was noise. This skill prevents that mistake.
+
+## Pricing
+
+Free: Basic P/C ratio for 3 tickers
+**Full bundle — $29 one-time**: Real/lottery separation + anomaly detection + per-expiry + unlimited tickers
+→ https://jaehyunpark.gumroad.com/l/tcyahy
+
+## Author
+
+Built from a real trading mistake that cost money. The real/lottery discovery is documented and battle-tested across 17 tickers over 2+ months.

package/bundled-skills/tokenwise/SKILL.md

@@ -0,0 +1,81 @@
+---
+name: tokenwise
+description: "Measurement-driven model router for Claude Code. Routes Haiku/Sonnet/Opus per task class, logs every routed task with real $ numbers, and A/B tests cheaper tiers before you trust the savings."
+category: developer-tools
+risk: safe
+source: community
+source_repo: CodeShuX/tokenwise
+source_type: community
+date_added: "2026-05-12"
+author: CodeShuX
+tags: [model-routing, token-optimization, cost-reduction, anthropic, haiku, sonnet, opus, claude-code, ab-testing, measurement]
+tools: [claude]
+license: "MIT"
+license_source: "https://github.com/CodeShuX/tokenwise/blob/main/LICENSE"
+---
+
+# TokenWise — Measurement-Driven Model Router
+
+## Overview
+
+A Claude Code skill that auto-routes subtasks to the cheapest model that can handle them (Haiku for grunt work, Sonnet for scoped reasoning, Opus only for synthesis), then logs every routed task to a local NDJSON with real token + cost numbers. Includes an A/B test subcommand that runs the same task across multiple tiers and scores quality, so the routing decisions are verified against the user's real workload — not estimated.
+
+Anthropic's own bug tracker (Issue #27665) reports 93.8% of Max-subscriber Claude Code tokens flow to Opus. Existing routers (claude-router, wshobson, VoltAgent) either pin models statically or route by vibes-based heuristics with no measurement. TokenWise fills the measurement gap.
+
+## When to use
+
+- Cutting Claude Code token spend without sacrificing output quality
+- Validating whether Haiku/Sonnet is "good enough" for a specific task class before trusting auto-routing
+- Auditing where Opus tokens are actually being burned
+- Logging per-session cost data for finance or chargeback
+
+## Subcommands
+
+- `/tokenwise:install` — guided installer with diff preview, automatic backups, and `--dry-run` mode
+- `/tokenwise:report` — per-session token + cost summary vs all-Opus baseline
+- `/tokenwise:summary [--week|--month|--all]` — historical aggregate with trend
+- `/tokenwise:ab "<task>"` — A/B test the same task at multiple tiers, generates a markdown comparison
+- `/tokenwise:undo` — restore CLAUDE.md / settings.json from backup
+
+## Routing taxonomy
+
+| Tier | Model | Task class |
+|---|---|---|
+| Mechanical | Haiku 4.5 | file reads, grep, format, rename, simple edits, doc lookups |
+| Scoped reasoning | Sonnet 4.6 | single-file refactor, scoped research, test writing |
+| Synthesis | Opus 4.7 | architecture decisions, multi-file refactor, security review |
+
+Safety caps:
+- Haiku never spawns further subagents
+- Max spawn depth = 2
+- Subagents that need a smarter model return to parent — they never escalate on their own
+- Tasks under 100 chars with no file context run inline (subagent overhead > savings)
+- Subagent context >30k tokens bumps a tier
+
+## Privacy
+
+Zero telemetry. All logs in `.tokenwise/log.ndjson` local to the project. Task descriptions truncated to 80 chars and stripped of file contents before logging. No analytics endpoint exists in the source.
+
+## Install
+
+In any Claude Code session:
+
+```
+/plugin marketplace add CodeShuX/tokenwise
+/plugin install tokenwise@tokenwise
+```
+
+Then run `/tokenwise:install` and follow the guided prompts.
+
+## Limitations
+
+- Token counts approximate to ±2% vs Anthropic billing
+- A/B test mode costs extra tokens (one task × N tiers) — intentional one-time validation
+- Anthropic-only by design (use LiteLLM or OpenRouter for cross-vendor)
+- Subagent `model:` param has known silent-fail bugs on some Claude Code builds — skill probes for this at install and refuses to configure if routing is broken
+
+## Source
+
+- Repo: https://github.com/CodeShuX/tokenwise
+- License: MIT
+- Author: CodeShuX

package/bundled-skills/wechat-official-account-strategist/SKILL.md

@@ -0,0 +1,124 @@
+---
+name: wechat-official-account-strategist
+description: "Grow WeChat Official Accounts (微信公众号) with high-conversion content strategy, title formulas, article architecture, and Mini-Program integration."
+category: marketing
+risk: safe
+source: community
+source_repo: demo112/yunqu-ai-skills
+source_type: community
+date_added: "2026-05-13"
+author: yundu-ai
+tags: [wechat, chinese-market, content-strategy, marketing, 公众号, 微信]
+tools: [claude, cursor, gemini]
+---
+
+# WeChat Official Account Strategist
+
+## Overview
+
+Expert strategist for WeChat Official Accounts (微信公众号), China's most powerful content marketing channel with 1.3 billion WeChat users. Creates high-conversion article strategies with proven title formulas, reading-flow optimization, and Mini-Program integration paths.
+
+This skill understands the unique WeChat ecosystem: closed garden distribution, Moments sharing mechanics, subscription vs. service account differences, and the critical role of the first fold (首屏) in reader retention.
+
+## When to Use This Skill
+
+- Use when creating articles for WeChat Official Accounts
+- Use when planning WeChat content strategy or editorial calendar
+- Use when optimizing article open rates and sharing rates
+- Use when designing WeChat-driven sales funnels
+- Use when converting readers to Mini-Program users or private traffic (私域流量)
+
+## How It Works
+
+### Step 1: Account Type Analysis
+
+Identify the account type and its constraints:
+- **Subscription Account (订阅号)**: 1 push per day, folded in subscription folder
+- **Service Account (服务号)**: 4 pushes per month, appears in main chat list
+- **Enterprise Account**: Internal communication and CRM
+
+### Step 2: Title Engineering
+
+Apply proven title formulas for WeChat:
+
+1. **Curiosity Gap**: "为什么XXX却YYY？" (Why X but Y?)
+2. **Counter-intuitive**: "一直以为XXX，原来YYY" (Always thought X, turns out Y)
+3. **Social Proof**: "XXX万人都在用的..." (X million people use this)
+4. **Urgency**: "再不看就晚了！" (Read before it's too late)
+5. **Value Promise**: "看完这篇，你就懂了..." (After reading this, you'll understand)
+6. **Identity**: "XXX的人，都有一共性" (People who X share this trait)
+
+### Step 3: Article Architecture
+
+Structure for WeChat's reading behavior:
+1. **First Fold (首屏)** - Hook + value promise (visible without scrolling)
+2. **Ramp (铺垫)** - Build context, establish credibility
+3. **Core Content (核心)** - Deliver the promised value
+4. **Emotional Peak (情感高潮)** - Create sharing motivation
+5. **CTA (行动呼唤)** - Clear next step (follow, share, click Mini-Program)
+
+### Step 4: Distribution Optimization
+
+Optimize for WeChat's sharing mechanics:
+- **Moments (朋友圈)**: Craft share-worthy pull quotes
+- **Direct Share (转发)**: Provide suggested forwarding text
+- **In-article search**: Place keywords for WeChat's article search index
+
+## Examples
+
+### Example 1: Tech Company Thought Leadership
+
+```
+Title: 程序员35岁危机？我和10个技术总监聊了聊，发现一个规律
+Structure:
+  首屏: 35岁真的会失业吗？数据说话...
+  铺垫: 调研背景，10位总监的行业分布
+  核心: 3个关键发现，打破刻板印象
+  高潮: "真正淘汰你的不是年龄，是..."
+  CTA: 关注公众号，回复"职场"获取完整报告
+```
+
+### Example 2: E-commerce Product Launch
+
+```
+Title: 用了这款面霜一个月，同事问我是不是做了医美
+Structure:
+  首屏: 真实使用对比图描述
+  铺垫: 皮肤困扰和选品过程
+  核心: 成分分析+使用感受+效果时间线
+  高潮: "最让我惊喜的是第三周..."
+  CTA: 点击小程序链接，限时优惠
+```
+
+## Best Practices
+
+- Write the first fold as if it is the only thing readers will see (many stop there)
+- Use short paragraphs (2-3 sentences) for mobile readability
+- Include 1 image every 300-500 words to break up text
+- End with a specific, low-friction CTA
+- Maintain consistent voice across articles to build brand recognition
+- Post at peak hours: 7-9am, 12-1pm, 8-10pm (China time)
+
+## Limitations
+
+- This skill generates text strategy; actual graphic design and layout require additional tools
+- WeChat algorithm updates may change optimal strategies
+- Industry-specific regulations (finance, health, education) may require compliance review
+
+## Security and Safety Notes
+
+- This skill generates content strategy and copy. It does not access WeChat APIs or accounts.
+- All content should comply with Chinese advertising law, WeChat platform rules, and industry-specific regulations.
+
+## Common Pitfalls
+
+- **Problem:** High open rate but low completion rate
+  **Solution:** Strengthen the first fold hook and reduce article length. WeChat readers have 3-5 minute attention windows.
+
+- **Problem:** Low sharing rate
+  **Solution:** Add an emotional peak before the CTA. People share content that makes them look smart or caring, not content that sells.
+
+## Related Skills
+
+- `xiaohongshu-content-strategist` - For short-form visual content on Xiaohongshu
+- `chinese-market-content-engineer` - For multi-platform Chinese content strategy

package/bundled-skills/x-twitter-scraper/SKILL.md

@@ -1,10 +1,10 @@
 ---
 name: x-twitter-scraper
-description: "X/Twitter automation skill for tweet search, follower export, media download, posting, replies, DMs, webhooks, MCP, SDKs, and the TweetClaw OpenClaw plugin."
+description: "X/Twitter automation skill for tweet search, follower export, posting, DMs, webhooks, MCP, SDKs, Hermes Tweet, and TweetClaw."
 category: data
 risk: safe
 source: community
-tags: "[twitter, x-api, tweet-search, twitter-api, twitter-scraper, follower-export, automation, mcp, sdk, webhooks, openclaw, tweetclaw]"
+tags: "[twitter, x-api, tweet-search, twitter-api, twitter-scraper, follower-export, automation, mcp, sdk, webhooks, hermes-agent, hermes-tweet, openclaw, tweetclaw]"
 date_added: "2026-02-28"
 ---
 
@@ -12,7 +12,7 @@ date_added: "2026-02-28"
 
 ## Overview
 
-Gives AI agents X (Twitter) data and automation workflows through the Xquik platform. Covers tweet search, advanced Twitter search, profile tweets, user lookup, follower export, media download, posting, replies, DMs, giveaway draws, account monitoring, webhooks, 23 bulk extraction tools, MCP, official SDKs, and the TweetClaw OpenClaw plugin.
+Gives AI agents X (Twitter) data and automation workflows through the Xquik platform. Covers tweet search, advanced Twitter search, profile tweets, user lookup, follower export, media download, posting, replies, DMs, giveaway draws, account monitoring, webhooks, 23 bulk extraction tools, MCP, official SDKs, the Hermes Tweet Hermes Agent plugin, and the TweetClaw OpenClaw plugin.
 
 ## When to Use This Skill
 
@@ -27,6 +27,7 @@ Gives AI agents X (Twitter) data and automation workflows through the Xquik plat
 - User wants to run a giveaway draw from tweet replies
 - User needs real-time monitoring of an X account (new tweets, follower changes)
 - User wants webhook delivery of monitored events
+- User wants the Hermes Tweet Hermes Agent plugin with `tweet_explore`, `tweet_read`, and approval-gated `tweet_action`
 - User wants the TweetClaw OpenClaw plugin instead of direct REST or MCP setup
 - User asks about trending topics on X
 
@@ -48,6 +49,16 @@ git clone https://github.com/Xquik-dev/x-twitter-scraper.git .claude/skills/x-tw
 git clone https://github.com/Xquik-dev/x-twitter-scraper.git .agents/skills/x-twitter-scraper
 ```
 
+### Use the Hermes Agent Plugin
+
+For Hermes Agent runtime tools, install Hermes Tweet. It wraps the same Xquik API with `tweet_explore` for endpoint discovery, `tweet_read` for read-only calls, and approval-gated `tweet_action` for writes and private actions.
+
+```bash
+hermes plugins install Xquik-dev/hermes-tweet --enable
+```
+
+Use Hermes Tweet when a Hermes Agent should search Twitter/X, read tweet replies, look up users, export followers, monitor tweets, post tweets, post replies, send DMs, or automate X actions with explicit approval gates.
+
 ### Use the OpenClaw Plugin
 
 For OpenClaw runtime tools, install TweetClaw. It wraps the same Xquik API with `explore` for endpoint discovery and `tweetclaw` for approved calls.
@@ -84,6 +95,7 @@ export XQUIK_API_KEY="xq_YOUR_KEY_HERE"
 | Write Actions | Send tweets, post replies, like, repost, follow, unfollow, and send DMs after explicit approval |
 | SDKs | Official TypeScript, Python, Ruby, Go, Kotlin, Java, PHP, C#, CLI, and Terraform clients |
 | MCP Server | StreamableHTTP endpoint for AI-native integrations |
+| Hermes Tweet Hermes Agent Plugin | Installable `hermes-tweet` runtime with `tweet_explore`, `tweet_read`, and approval-gated `tweet_action` tools |
 | TweetClaw OpenClaw Plugin | Installable `@xquik/tweetclaw` runtime with `explore` and `tweetclaw` tools |
 
 ## Examples
@@ -113,6 +125,11 @@ export XQUIK_API_KEY="xq_YOUR_KEY_HERE"
 "Monitor @openai for new tweets and notify me via webhook"
 ```
 
+**Use Hermes Agent:**
+```
+"Use Hermes Tweet to search Twitter/X for this launch, read the tweet replies, and prepare a draft reply for approval"
+```
+
 **Bulk extraction:**
 ```
 "Extract all followers of @anthropic"
@@ -150,6 +167,8 @@ export XQUIK_API_KEY="xq_YOUR_KEY_HERE"
 
 https://github.com/Xquik-dev/x-twitter-scraper
 
+Hermes Tweet Hermes Agent plugin: https://github.com/Xquik-dev/hermes-tweet
+
 TweetClaw OpenClaw plugin: https://github.com/Xquik-dev/tweetclaw
 
 **Maintained By:** [Xquik](https://xquik.com)

package/bundled-skills/xiaohongshu-content-strategist/SKILL.md

@@ -0,0 +1,117 @@
+---
+name: xiaohongshu-content-strategist
+description: "Create viral Xiaohongshu (小红书) content with platform-native strategy, save-rate optimization, trending formats, and search SEO for China's #1 lifestyle platform."
+category: marketing
+risk: safe
+source: community
+source_repo: demo112/yunqu-ai-skills
+source_type: community
+date_added: "2026-05-13"
+author: yundu-ai
+tags: [xiaohongshu, chinese-market, content-strategy, social-media, marketing, 红书, 小红书]
+tools: [claude, cursor, gemini]
+---
+
+# Xiaohongshu Content Strategist
+
+## Overview
+
+Expert content strategist for Xiaohongshu (小红书), China's most influential lifestyle and shopping platform with 300M+ monthly active users. Creates platform-native content optimized for the unique Xiaohongshu algorithm, which prioritizes saves over likes. Bilingual Chinese/English output with cultural sensitivity.
+
+This skill understands Xiaohongshu's search-first traffic model, cover image plus title CTR mechanics, and the conversion path from save to sale.
+
+## When to Use This Skill
+
+- Use when creating content for Xiaohongshu
+- Use when optimizing existing content for better Xiaohongshu performance
+- Use when planning a Xiaohongshu content calendar or strategy
+- Use when adapting international brand content for the Chinese market via Xiaohongshu
+- Use when analyzing Xiaohongshu competitors
+
+## How It Works
+
+### Step 1: Analyze the Topic
+
+Understand the target audience, product, or message. Identify primary keywords for Xiaohongshu's search algorithm. Research trending formats in the relevant category.
+
+### Step 2: Choose the Content Format
+
+Select from proven formats based on the topic:
+
+| Format | Best For | Example |
+|--------|----------|---------|
+| Before/After | Transformations |妆前妆后、装修前后 |
+| Step-by-Step | Tutorials | 5步学会xxx |
+| Comparison | Decisions | A vs B 实测 |
+| Hidden Gems | Discovery | 被低估的xxx |
+| List/Rankings | Quick value | 2025必买的10件 |
+
+### Step 3: Generate the Content Package
+
+For each post, provide:
+1. **Cover Image Brief** - Visual concept, text overlay under 10 chars, color mood
+2. **Title** (2-3 options) - Primary keyword in first 8 chars, emotional trigger, 18-22 chars optimal
+3. **Body Content** - Hook sentence, short paragraphs, strategic emoji, highlighted key info, CTA
+4. **Hashtags** - 3-5 mix of high-volume and niche tags
+5. **Comment Engagement Plan** - Seed comments and anticipated Q&A
+
+### Step 4: Optimize for the Algorithm
+
+Apply these ranking factors in priority order:
+1. Save rate - number one ranking signal, content must be reference-worthy
+2. Click-through rate - driven by cover image plus title
+3. Comment depth - conversation quality over count
+4. Completion rate - users who read to the end
+
+## Examples
+
+### Example 1: Beauty Product Review
+
+Title: 用了28天，皮肤真的变好了｜实测这款平价精华
+Cover: Before/After face photo with product in corner, soft pink overlay
+Body: 这款精华我用了整整28天，今天来交作业...
+Hashtags: #平价护肤 #精华推荐 #28天打卡
+Seed Comment: "姐妹们，我油皮可以用吗？"
+
+### Example 2: Travel Destination
+
+Title: 上海被低估的咖啡馆！拍照绝了
+Cover: Cafe interior shot with warm tones, location pin overlay
+Body: 周末不想人挤人？这家藏在法租界的小店...
+Hashtags: #上海咖啡 #周末去哪 #小众探店
+Seed Comment: "地址在哪里呀？"
+
+## Best Practices
+
+- Sound like a real person sharing a discovery, not a brand broadcasting
+- Front-load keywords in titles (first 8 characters)
+- Use numbers and specific results in titles
+- Keep paragraphs to 2-3 sentences max
+- Include a clear save-worthy takeaway
+- Do not use corporate marketing language
+- Do not ignore mobile formatting (most users are on phones)
+- Do not post without relevant hashtags
+
+## Limitations
+
+- This skill generates text content strategy; actual image/video creation requires additional tools
+- Trending topics and algorithm details may shift; always validate with current platform data
+- Cultural nuances in specific sub-communities may require human review
+
+## Security and Safety Notes
+
+- This skill generates content strategy and copy. It does not access Xiaohongshu APIs or user accounts.
+- All content should comply with Chinese advertising law and platform community guidelines.
+
+## Common Pitfalls
+
+- **Problem:** Low engagement despite good content
+  **Solution:** Check title CTR - the cover image plus title combo drives 80 percent of click-through. A/B test 2-3 title options.
+
+- **Problem:** Content gets flagged or removed
+  **Solution:** Avoid absolute claims and ensure product reviews disclose sponsorships per platform rules.
+
+## Related Skills
+
+- `wechat-official-account-strategist` - For long-form content strategy on WeChat
+- `chinese-market-content-engineer` - For multi-platform Chinese content strategy

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "opencode-skills-collection",
-  "version": "3.0.13",
+  "version": "3.0.15",
   "description": "OpenCode CLI plugin that automatically downloads and keeps skills up to date.",
   "main": "dist/index.js",
   "types": "dist/index.d.ts",

package/skills_index.json

@@ -18719,6 +18719,28 @@
       "reasons": []
     }
   },
+  {
+    "id": "mcp-tool-developer",
+    "path": "skills/mcp-tool-developer",
+    "category": "developer-tools",
+    "name": "mcp-tool-developer",
+    "description": "Build Model Context Protocol (MCP) servers and tools from scratch. Full-stack MCP development with TypeScript/Python, testing, deployment, and registry publishing.",
+    "risk": "safe",
+    "source": "community",
+    "date_added": "2026-05-13",
+    "plugin": {
+      "targets": {
+        "codex": "supported",
+        "claude": "supported"
+      },
+      "setup": {
+        "type": "none",
+        "summary": "",
+        "docs": null
+      },
+      "reasons": []
+    }
+  },
   {
     "id": "memory-forensics",
     "path": "skills/memory-forensics",
@@ -20087,6 +20109,28 @@
       "reasons": []
     }
   },
+  {
+    "id": "news-sentiment-engine",
+    "path": "skills/news-sentiment-engine",
+    "category": "research",
+    "name": "news-sentiment-engine",
+    "description": "Multi-source RSS news aggregation with Claude-powered sentiment analysis and structured briefing output",
+    "risk": "safe",
+    "source": "community",
+    "date_added": "2026-05-13",
+    "plugin": {
+      "targets": {
+        "codex": "supported",
+        "claude": "supported"
+      },
+      "setup": {
+        "type": "none",
+        "summary": "",
+        "docs": null
+      },
+      "reasons": []
+    }
+  },
   {
     "id": "nextjs-app-router-patterns",
     "path": "skills/nextjs-app-router-patterns",
@@ -21256,6 +21300,28 @@
       "reasons": []
     }
   },
+  {
+    "id": "options-flow-analyzer",
+    "path": "skills/options-flow-analyzer",
+    "category": "finance",
+    "name": "options-flow-analyzer",
+    "description": "Real vs lottery call separation for options P/C ratio analysis \u2014 prevents signal inversion from deep OTM noise",
+    "risk": "safe",
+    "source": "community",
+    "date_added": "2026-05-13",
+    "plugin": {
+      "targets": {
+        "codex": "supported",
+        "claude": "supported"
+      },
+      "setup": {
+        "type": "none",
+        "summary": "",
+        "docs": null
+      },
+      "reasons": []
+    }
+  },
   {
     "id": "oral-health-analyzer",
     "path": "skills/oral-health-analyzer",
@@ -29104,6 +29170,28 @@
       "reasons": []
     }
   },
+  {
+    "id": "tokenwise",
+    "path": "skills/tokenwise",
+    "category": "developer-tools",
+    "name": "tokenwise",
+    "description": "Measurement-driven model router for Claude Code. Routes Haiku/Sonnet/Opus per task class, logs every routed task with real $ numbers, and A/B tests cheaper tiers before you trust the savings.",
+    "risk": "safe",
+    "source": "community",
+    "date_added": "2026-05-12",
+    "plugin": {
+      "targets": {
+        "codex": "supported",
+        "claude": "supported"
+      },
+      "setup": {
+        "type": "none",
+        "summary": "",
+        "docs": null
+      },
+      "reasons": []
+    }
+  },
   {
     "id": "tool-design",
     "path": "skills/tool-design",
@@ -30978,6 +31066,28 @@
       "reasons": []
     }
   },
+  {
+    "id": "wechat-official-account-strategist",
+    "path": "skills/wechat-official-account-strategist",
+    "category": "marketing",
+    "name": "wechat-official-account-strategist",
+    "description": "Grow WeChat Official Accounts (\u5fae\u4fe1\u516c\u4f17\u53f7) with high-conversion content strategy, title formulas, article architecture, and Mini-Program integration.",
+    "risk": "safe",
+    "source": "community",
+    "date_added": "2026-05-13",
+    "plugin": {
+      "targets": {
+        "codex": "supported",
+        "claude": "supported"
+      },
+      "setup": {
+        "type": "none",
+        "summary": "",
+        "docs": null
+      },
+      "reasons": []
+    }
+  },
   {
     "id": "weightloss-analyzer",
     "path": "skills/weightloss-analyzer",
@@ -31599,7 +31709,7 @@
     "path": "skills/x-twitter-scraper",
     "category": "data",
     "name": "x-twitter-scraper",
-    "description": "X/Twitter automation skill for tweet search, follower export, media download, posting, replies, DMs, webhooks, MCP, SDKs, and the TweetClaw OpenClaw plugin.",
+    "description": "X/Twitter automation skill for tweet search, follower export, posting, DMs, webhooks, MCP, SDKs, Hermes Tweet, and TweetClaw.",
     "risk": "safe",
     "source": "community",
     "date_added": "2026-02-28",
@@ -31638,6 +31748,28 @@
       "reasons": []
     }
   },
+  {
+    "id": "xiaohongshu-content-strategist",
+    "path": "skills/xiaohongshu-content-strategist",
+    "category": "marketing",
+    "name": "xiaohongshu-content-strategist",
+    "description": "Create viral Xiaohongshu (\u5c0f\u7ea2\u4e66) content with platform-native strategy, save-rate optimization, trending formats, and search SEO for China's #1 lifestyle platform.",
+    "risk": "safe",
+    "source": "community",
+    "date_added": "2026-05-13",
+    "plugin": {
+      "targets": {
+        "codex": "supported",
+        "claude": "supported"
+      },
+      "setup": {
+        "type": "none",
+        "summary": "",
+        "docs": null
+      },
+      "reasons": []
+    }
+  },
   {
     "id": "xlsx-official",
     "path": "skills/xlsx-official",