/nav-multi
DEPRECATED. Multi-Claude shell orchestration is superseded by native Claude Code Workflows. Auto-invokes on "run multi-agent workflow", "parallel agents for", "multi-claude for", "spawn agents for" to redirect the user to the native Workflow and Agent tools.
$ npx -y skills add alekspetrov/navigator --skill nav-multi --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
- Slash command
/nav-multi
Context preview
The summary Claude sees to decide when to auto-load this skill.
DEPRECATED. Multi-Claude shell orchestration is superseded by native Claude Code Workflows. Auto-invokes on "run multi-agent workflow", "parallel agents for", "multi-claude for", "spawn agents for" to redirect the user to the native Workflow and Agent tools.
SKILL.md
nav-multi.SKILL.mdname: nav-multi
description: DEPRECATED. Multi-Claude shell orchestration is superseded by native Claude Code Workflows. Auto-invokes on "run multi-agent workflow", "parallel agents for", "multi-claude for", "spawn agents for" to redirect the user to the native Workflow and Agent tools.
allowed-tools: Read
version: 2.0.0
nav-multi (DEPRECATED)
**Deprecated as of 2026-06 (TASK-25).**
Navigator's marker-coordinated, headless `claude -p` multi-agent scripts are superseded by Claude Code's **native orchestration** — more reliable, no external scripts, and what this project itself now uses for its own audits and planning:
- **`/workflows` — Dynamic Workflows**: deterministic multi-phase orchestration
(plan → implement → test → review), parallel/pipeline fan-out, structured output, in-session resume. No marker files, no context bloat.
- **The Agent tool / subagents**: parallel independent agents with optional git
worktree isolation; can run in the background and report back.
When invoked
Tell the user the multi-Claude scripts are deprecated and point them to the native tools above. **Do NOT run** `navigator-multi-claude*.sh`.
Example response:
> Multi-Claude shell workflows are deprecated. Use native Claude Code > orchestration instead: `/workflows` for multi-phase work, or the Agent tool > for parallel subagents. Both supersede the old marker-file scripts.
History
The scripts (`scripts/navigator-multi-claude*.sh`, `sub-claude-monitor.sh`, `resume-workflow.sh`, `multi-claude-dashboard.sh`) remain in the repo for reference under a `DEPRECATED` banner but are unmaintained (~30% success rate). See `.agent/tasks/TASK-25-fix-multi-claude-reliability.md` for the deprecation rationale and the research that informed it.
Read more
name: nav-multi description: DEPRECATED. Multi-Claude shell orchestration is superseded by native Claude Code Workflows. Auto-invokes on "run multi-agent workflow", "parallel agents for", "multi-claude for", "spawn agents for" to redirect the user to the native Workflow and Agent tools. allowed-tools: Read version: 2.0.0
nav-multi (DEPRECATED)
**Deprecated as of 2026-06 (TASK-25).**
Navigator's marker-coordinated, headless `claude -p` multi-agent scripts are superseded by Claude Code's **native orchestration** — more reliable, no external scripts, and what this project itself now uses for its own audits and planning:
- **`/workflows` — Dynamic Workflows**: deterministic multi-phase orchestration
(plan → implement → test → review), parallel/pipeline fan-out, structured output, in-session resume. No marker files, no context bloat.
- **The Agent tool / subagents**: parallel independent agents with optional git
worktree isolation; can run in the background and report back.
When invoked
Tell the user the multi-Claude scripts are deprecated and point them to the native tools above. **Do NOT run** `navigator-multi-claude*.sh`.
Example response:
> Multi-Claude shell workflows are deprecated. Use native Claude Code > orchestration instead: `/workflows` for multi-phase work, or the Agent tool > for parallel subagents. Both supersede the old marker-file scripts.
History
The scripts (`scripts/navigator-multi-claude*.sh`, `sub-claude-monitor.sh`, `resume-workflow.sh`, `multi-claude-dashboard.sh`) remain in the repo for reference under a `DEPRECATED` banner but are unmaintained (~30% success rate). See `.agent/tasks/TASK-25-fix-multi-claude-reliability.md` for the deprecation rationale and the research that informed it.
Finish What You Start Sessions that last. AI that learns. Features that ship.
Repo: alekspetrov/navigator
Other skills on navigator.
- /backend-endpoint
Create REST/GraphQL API endpoint with validation, error handling, and tests. Auto-invoke when user says "add endpoint", "create API", "new route", or "add route".
Open skill - /backend-test
Generate backend tests (unit, integration, mocks) for existing code. Auto-invoke when user says "write test for", "add test", "test this", or "create test".
Open skill - /database-migration
Create database migration with schema changes and rollback. Auto-invoke when user says "create migration", "add table", "modify schema", or "change database".
Open skill - /frontend-component
Create React/Vue component with TypeScript, tests, and styles. Auto-invoke when user says "create component", "add component", "new component", or "build component".
Open skill - /frontend-test
Generate frontend component tests (React Testing Library, Vue Test Utils, snapshot) for existing components. Auto-invoke when user says "test this component", "write component test", or "add component test".
Open skill - /nav-brief
Render a one-screen intent brief (Goal/Scope/Approach/Limits/Verify/Won't-do) before implementing ambiguous task-shaped prompts, triggered by the nav_brief.py UserPromptSubmit hook. Confirms scope with max 2 open questions before touching files; detects brief drift mid-task.
Open skill