agents

Routes adaptive multi-agent deliberation with fractal context cycles. Use when using /cdo, think/deep/debug/parliament work, or long runs paired with autoresearch scheduling.

Use when you need an evidence-first provenance report from `cass`, repo docs, `.lev`, `~/.agents/diagrams`, `qmd`, or Grep/Glob tools.

Use the codebase knowledge graph for structural code queries. Triggers on: explore the codebase, understand the architecture, what functions exist, show me the structure, who calls this function, what does X call, trace the call chain, find callers of, show dependencies, impact analysis, dead code, unused functions, high fan-out, refactor candidates, code quality audit, graph query syntax, Cypher query examples, edge types, how to use search_graph.

Handle blocked destructive commands. Use when dcg blocks rm -rf, git reset --hard, DROP DATABASE, kubectl delete, or when configuring agent safety guardrails.

Use when running a visible Lev autopilot that coordinates research, interview, arch, propose, AgentPing, warm-worker intent, realtime dashboard intent, and lev exec updates across the lifecycle.

Heartbeat-driven autonomous development loop using lev loop autodev, with interval, budget, and tick controls for in-process SDLC execution.

Mine atuin history and shell logs to find repetitive workflows, build Rust CLI or bash automations. Use when analyzing command patterns or finding automation opportunities.

Use when the user wants the current conversation or captured context turned into a lifecycle artifact without a long interview, especially before proposal planning.

Use when installing, configuring, scripting, troubleshooting, or operating the Coolify CLI for Coolify Cloud or self-hosted Coolify instances.

Multi-perspective project analysis via NTM reasoning-mode agent swarm. Use when "modes of reasoning", "multi-perspective analysis", or "epistemological review" of any project.

Reminder to use rp-cli

Use when executing implementation plans with independent tasks - dispatch coder agents for bounded slices and run reviewer agents on cadence, size, risk triggers, and final integration

Deterministic lev.now renders JSON specs to styled HTML with publish and attach. Use when user wants dashboards, reports, timelines, or rich pages instead of large markdown tables.

Use when reconciling chezmoi local-vs-remote drift, reviewing pending machine-state changes, or performing safe remote-first dotfiles sync across machines.

Use when dispatching Codex CLI (`codex exec`) as a bounded background worker with subagent ergonomics — fire a scoped prompt, harvest the final message, verify independently before reporting.

lev eval, core/eval, design eval suites, runtime scenarios, acceptance criteria, adversarial probes, or proof gates for agentic features before implementation or promotion.

Use when guiding decision interviews that produce a design artifact in .lev/pm/designs before proposal or implementation planning.

Use when compiling an aligned design, capture, or chat intent into execution-ready task DNA and execution YAML.

Use when any research, search, or information gathering is needed.

[OMX] Adversarial dynamic e2e QA workflow - generate hostile scenarios, test, verify, fix, report, and clean up

Use when routing tracked work through lifecycle lanes, resolving workstream context, or deciding which lifecycle skill owns the next entity transition.

Use when inventorying conversation state into durable artifacts, routing shovel-ready items toward exec, or deep-processing a brain dump.

Use when running execution-ready task slices, dispatching verified `lev exec` flows, or reviewing implementation work in the Exec lane.

Config-driven workshop intake for URLs, repos, media, PDFs, and skill packages. Use when user wants intake, download, or external content analyzed with project playbooks, ending with a brief, routed recommendation, validation plan, and action dashboard.

Routes UX/product design runs, user research/persona synthesis, wireframes, flows, IA, pencil visuals, and agentic CLI UX. Infers full run vs research vs CDO-backed persona mode from context, asking one question only when ambiguous.

Use when converting a user request into a Codex goal that will use `/exec`, Lev exec, or Lev Ralph as bounded execution tools.

Thin Notion API CLI for doctor, search, queries, pages, and comments. Use when changing notionctl/lev-notionctl or keeping triage and policy out of the CLI surface.

Checkpoint sync for the current checkout and recursive dirty submodules. In feature branches, paired main checkout and feature->main landing require explicit --main.

Use when the user asks to prove, spike, prototype, POC, validate a production path, compare landing zones, or decide whether an idea should move to propose, monitor, reject, or archive.

Use when verified work is ready for QA, acceptance, checkpointing, learning, recommendation, and handoff.

Use when a long brainstorm or brain dump needs deep capture, prior-art routing, and durable next states.

Use when creating skills, converting docs/repos/PDFs to skills, installing external skills, auditing skill security, or merging skills

Filesystem-first local skill lookup and ranking from lev-skills inventory. Use when work, lev, or hubs need to discover installed skills without low-quality fallbacks.

Use when invoking, debugging, or dogfooding the Lev CLI; choosing between work, exec, stack, FlowMind, validation, traces, receipts, or runtime health commands.

Use when listing, inspecting, running, or managing prompt-stack sessions. Triggers on "stack", "list stacks", "show stack", "init stack", "run stack", "stack status".

Build a 3-week coding-agent capture dashboard. Cross-references cass session inventory with .lev/pm/ workstream/handoff citations, detects resume-chain lineage, classifies orphan chains against the workstream registry, emits a canonical coverage.json + thin static viewer. Use when auditing how much agent work is captured by Lev workstreams, finding orphan transcripts, building a healing pipeline, or scheduling a weekly coverage report.

Use when closing or pausing a session by updating workstream state and copying a resume prompt.

Use when scanning, finding, resuming, merging, branching, or untangling Lev workstreams.

Named Tmux Manager for orchestrating Claude, Codex, and Gemini agents in tmux sessions. Use when you need to spawn, attach, message, inspect, or doctor a multi-agent tmux workspace.

Project alignment validator. Loads .lev/validation-gates.yaml and checks current work against all applicable gates. Reports pass/fail/aspirational per gate. Integrates with /work ALIGN phase.