core

Implement or review the experimental WorkOS auth.md v0.6.0 agent-identity proposal. Use for "implement auth.md", "add agent registration", "design delegated agent signup", "support service_auth", "mint or verify an ID-JAG", "build an agent identity provider", or agent-facing OAuth consent, claims, and revocation. Distinct from Better Auth Agent Auth Protocol, OAuth Dynamic Client Registration, and RFC 8628 Device Authorization.

Compare installed plugin versions against GitHub to confirm skills and agents are current. Use at session start, for "is everything up to date", "are my plugins up to date", "are our plugins current", "check for plugin updates", "is anything stale", "am I on the latest version", "did that plugin update land", or when a skill behaves differently than documented or references features you do not recognize. Checks any plugin in the marketplace, not just core. Completes in under 100ms per plugin.

Generate and install Codex agent adapters so plugin personas are invocable from the Codex CLI. Use for "set up codex agents", "install the codex adapters", "make these agents work in Codex", or when a persona is missing from a Codex session.

Proactive self-audit covering incomplete changes, untested assumptions, pattern violations, buried concerns, and cleanup debt. Invoke before ending a session, marking a task done, or saying "complete", and when the user asks "are you sure?", "did you miss anything?", "anything else?", or "audit your work". Do not wait to be asked.

Route to the right agent, skill, or service provider. Use for "who handles X?", "what agents are available?", "team roster", "what skills are available?", "where do I find skill X?", or "who should I talk to about Z?". Also covers agent-adapter availability, installing skills, and org email drafting. Route SOC 2 and audit readiness to Anthony in product-skills.

This skill should be used when the user mentions a behavioral rule they want enforced, says "always do X", "never do Y", "stop doing Z", "from now on", asks about HammerTime rules, wants to create a stop hook rule, mentions behavioral guardrails, or wants to understand how the HammerTime stop hook system works. Teaches how to write rules for the HammerTime stop hook system.

List, enable, or disable individual core hooks and check their prerequisites. Use for "list hooks", "enable a hook", "disable a hook", "hook setup", "hooks config", or a [BOPEN-HOOKS-SETUP] directive. For the full harness view, use setup.

This skill should be used for human-facing prose — emails, docs, reports, posts, release notes, and commit messages — when the user asks to "humanize", "make this sound less AI", "de-AI this", "this sounds like ChatGPT", or "edit this". Preserve facts, evidence, citations, house style, and intended meaning while removing clustered AI-writing patterns.

This skill should be used when the user wants to plan a project, feature, or bug fix using Linear as the agent control plane. Use when the user says "plan this in Linear", "create Linear tickets", "break this down into tasks", "push to Linear", "set up our board for this feature", or wants to turn a description or spec into well-structured, agent-ready Linear issues. Requires the official Linear MCP server to be configured.

Inject compressed SKILL-MAP and AGENT-MAP directive blocks into CLAUDE.md so skill names and agent IDs survive a long session. Use when modifying a CLAUDE.md, when adding a skill or agent to a plugin, or for "update the skill map", "register this agent", "skills keep getting forgotten", "sync skills to CLAUDE.md".

Search and recall previous Claude Code sessions. Use for "remember when we...", "what did we do about...", "find that conversation where...", "when did I last work on...", "what was that command we used for...", or any reference to prior work absent from the current session, including vague ones like "that thing we did last week".

This skill should be used when the user asks "what runtime am I in", "detect the execution environment", "adapt behavior for Claude Code vs a hosted bot", or "check what tools are available in this environment". Detects the current agent execution environment (Claude Code, Vercel Sandbox, or local dev) and adapts behavior accordingly — for agents or bots that run in Claude Code as subagents AND as hosted bots, or a SOUL.md/SKILL.md that must work across runtimes.

Audit which bOpen plugins, CLIs, env keys, third-party skills, agents, and hooks are installed across the harness, then emit a runtime-tailored instruction plan; it installs nothing itself. Use for "bopen setup", "setup ui", "harness install", "audit my setup", "install everything", or "unified installer". For a single hook, use hook-manager.

Generate a navigable visual map of a codebase or system for orientation. Use when the user asks to "map this codebase", "show me how this fits together", "give me a tour", or needs a visual orientation artifact before working in unfamiliar code.