onsails

Code & Development Listed

This skill should be used when working with Rust code, reviewing Rust code, managing Rust dependencies, creating Rust projects, or fixing Rust compilation errors. It provides strict coding standards (especially FAIL FAST error handling), workspace architecture guidance, dependency management automation, and common Rust patterns.

25 Updated today

right-cron

Manages cron jobs for this Right Agent agent via MCP tools. Creates, updates, and deletes cron specs stored in the agent database. The Rust runtime handles scheduling and execution automatically. Use when the user mentions cron jobs, scheduled tasks, reminders, one-shot tasks, or recurring tasks.

right-learn-skill

Use ONLY when the user explicitly asks you to save / remember / fix a reusable workflow. Routine learning is handled automatically by the platform's probe-writer; do not invoke this skill based on your own judgment that a workflow might be reusable.

right-mcp

Finds and adds MCP servers for this Right Agent agent. Searches for OAuth-capable endpoints first (Claude Code / Codex integration docs), falls back to API-key endpoints. All management goes through the user's Telegram dashboard MCP view — the agent never handles credentials directly. Use when the user asks to add, connect, or set up an MCP server or integration.

right-skills

Manages Agent Skills for this Right Agent agent via skills.sh (Vercel's Agent Skills directory). Searches the skills.sh registry, installs skills by owner/repo slug, removes installed skills, and lists all installed skills. Use when the user wants to find, install, remove, update, or list Claude Code skills, or mentions skills.sh, skill packages, or skill management.

right-composio

Use BEFORE calling any mcp__right__composio__* tool — especially COMPOSIO_SEARCH_TOOLS, COMPOSIO_MULTI_EXECUTE_TOOL, or COMPOSIO_REMOTE_WORKBENCH — and whenever a request involves Notion, Gmail, Google Calendar, Google Drive, Slack, GitHub, Linear, or any other service routed through Composio. Sessions that skip this skill burn 300-500K context tokens on repeated tool searches and inline payloads ($200+ per session observed in cache-write cost). Read once per session, on resume, and any turn that mentions a Composio-fronted service.

right-reflect

Inspects this agent's own past reasoning by reading the conversation-history JSONL files Claude Code writes inside the sandbox. Use when the user asks "why did you...", "what were you thinking when...", or to debug a wrong decision the agent made in an earlier turn or session. Reads thinking blocks, tool calls, and tool results to reconstruct prior reasoning. Filesystem-only — no MCP calls, no DB.

right-memory

Manage your long-term memory powered by Hindsight