Valeriy Sakharov
Community@valeratrades · France, Clermont-Ferrand
Quant
Agent Skills by Valeriy Sakharov
Showing 7 vetted skills indexed across 1 GitHub repositories.
valeratradesCommand Development
Create Claude Code slash commands with frontmatter, arguments, and bash execution.
valeratradesSkill Development
Create Claude Code plugin skills with progressive disclosure and trigger-based descriptions.
valeratradesPlugin Settings
Implements per-project plugin configuration using .claude/plugin-name.local.md files with YAML frontmatter.
valeratradesPlugin Structure
Organize Claude Code plugin directories, manifests, and components using auto-discovery conventions.
valeratradesHook Development
Create and validate event-driven hooks for Claude Code plugin automation.
valeratradesMCP Integration
Configure Model Context Protocol servers in Claude Code plugins for external service integration.
valeratradesAgent Development
Create autonomous agent configurations with frontmatter, system prompts, and triggering examples for Claude Code plugins.
Frequently Asked Questions About Valeriy Sakharov
FAQPage SchemaWhat tasks can I accomplish using Valeriy Sakharov's skills?▼
You can build complete Claude Code plugins: scaffold plugin structure with plugin.json, create slash commands with dynamic arguments, implement PreToolUse/PostToolUse/Stop hooks, author subagents with system prompts, integrate MCP servers, and store per-project configuration in .local.md files.
Who are these Claude Code plugin skills designed for?▼
These skills target developers extending Claude Code with custom plugins. They serve engineers who need guidance on plugin architecture, command frontmatter, hook events like SessionStart and UserPromptSubmit, agent triggering conditions, and Model Context Protocol server configuration.
How do the hook development capabilities work?▼
The hook-development skill guides creation of event-driven hooks for PreToolUse, PostToolUse, Stop, SubagentStop, SessionStart, SessionEnd, UserPromptSubmit, PreCompact, and Notification events. It covers prompt-based hooks, tool-use validation, blocking dangerous commands, and ${CLAUDE_PLUGIN_ROOT} usage.
What MCP integration options are documented?▼
The mcp-integration skill documents adding Model Context Protocol servers to Claude Code plugins via .mcp.json configuration. It covers all four server transport types: SSE, stdio, HTTP, and WebSocket, enabling plugins to connect external services and data sources.
How are plugin settings and configuration stored?▼
The plugin-settings skill documents the .claude/plugin-name.local.md pattern for user-configurable plugins. Configuration is stored as YAML frontmatter with markdown content, enabling per-project plugin state files that make plugin behavior configurable without code changes.