Agent Claude Code
Reviews architectural decisions for Genesis. Use when evaluating new subsystems, integration patterns, or significant refactors. Enforces Genesis design principles and catches long-term liabilities.
Agent Claude Code
Maps project structure, identifies patterns and conventions, traces architecture layers, and documents dependencies. Use when onboarding to a codebase, auditing structure, or before major refactors.
Agent Claude Code
Diagnoses Genesis subsystem failures. Use when something is broken, degraded, or reporting unexpected state. Knows the full observability stack, event schema, and where to look for root causes.
Agent Claude Code
Deep research agent with full web + code intelligence. Use for any task requiring web fetching, searching, codebase exploration, or multi-source synthesis. Prefer this over generic Explore agents for research tasks.
Agent Claude Code
Security-reviews Genesis code changes. Use for diffs touching auth, credentials/secrets, financial transactions, autonomy approval gates, external input handling (Telegram/dashboard/MCP), SQL, subprocess, or path handling. Reports findings in CRITICAL/HIGH/LOW tiers.
Command Claude Code
Activate Chrome DevTools MCP for heavy browser sessions (network inspection, Lighthouse, performance tracing). Deactivate when done.
Command Claude Code
Review all changes in the current working tree (staged and unstaged).
Command Claude Code
Read the specified design doc section.
Command Claude Code
Run a fresh-context, un-anchored ADVERSARIAL-LOGIC review of the current diff before pushing — the review that catches real bugs, not a lint/secrets/"structural CLEAN" scan. Use it on any substantial change, and ALWAYS before pushing a hook, gate, parser, or security-surface change.
Command Claude Code
Read the Genesis build phases document: docs/architecture/genesis-v3-build-phases.md.
Command Claude Code
Evaluate technologies, tools, articles, videos, and competitive developments against Genesis architecture. Invoke when reviewing external sources, conducting competitive analysis, evaluating tools for adoption, or during strategic reflection on industry developments.
Command Claude Code
Toggle Genesis context for interactive Claude Code sessions.
Command Claude Code
List the Genesis skill catalog (Tier-1 always-indexed + Tier-2 specialized) with a one-line description each, so you can discover under-used capabilities on demand. Optional keyword filter. (Does not list Claude Code plugin skills.).
Command Claude Code
First-run onboarding or config repair. Routes on two signals — whether the Layer-1 install finished, and whether onboarding has completed (/.genesis/setup-complete). A broken/half-finished install is repaired; a complete-but-unconfigured install is onboarded; a configured install is refreshed.
Command Claude Code
Shelve the current session — create a bookmark so you can find and resume it later with /unshelve.
Command Claude Code
Guided task intake — interviews the user, writes a plan following TASKINTAKE.md format, validates structure, then submits to the autonomous task executor.
Command Claude Code
Search for shelved sessions — find past bookmarked sessions by keyword or browse recent ones.
Command Claude Code
Evaluate content through the lens of what Genesis knows about the user. Articles, research, ideas, tools, or any content the user wants analyzed for personal relevance and actionable value. The differentiator is the user model — Genesis's accumulated understanding of who this person is.
Hook Claude Code
Runs when a session starts on startup, resume, clear, startup, resume, startup, resume and clear, executing genesis_session_context.py, check_stale_pending.py, surface_pr_updates.py, surface_open_prs.py and cbm-session-reminder.sh (5 commands). From WingedGuardian/GENesis-AGI.
Hook Claude Code
Runs when you submit a prompt, before the agent sees it, executing genesis_urgent_alerts.py, review_enforcement_prompt.py, proactive_memory_hook.py, skill_injection_hook.py, contribution_offer_hook.py and session_activity_touch.sh (6 commands). From WingedGuardian/GENesis-AGI.
Hook Claude Code
Runs before the agent uses a tool for Read, Grep, Glob, Bash, Bash, Bash, Bash, Bash, Bash, Bash, Bash, Bash, Bash, Bash, Bash, Bash, WebFetch, Write, Edit, WebFetch, WebSearch, Agent, ExitWorktree and EnterWorktree tool calls, executing procedure_advisor.py, cbm-discovery-gate.sh, background_pipe_guard.py…
Hook Claude Code
Runs when the agent finishes a response, executing genesis_stop_hook.py and deliverable_gate_guard.py (2 commands). From WingedGuardian/GENesis-AGI.
Hook Claude Code
Runs when a session ends, executing genesis_session_end.py. From WingedGuardian/GENesis-AGI.
Hook Claude Code
Runs before the context is compacted, executing genesis_precompact.py. From WingedGuardian/GENesis-AGI.