local-dev-diagnostics

Diagnose Multica dev stack health, ports, Next.js profiling, and MCP smoke checks.

Updated Jun 8, 2026
One-click install
npx skills add https://github.com/vovanostm-public/multica --skill local-dev-diagnostics-vovanostm-public
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: local-dev-diagnostics
Source: https://github.com/vovanostm-public/multica/tree/main/.agents/skills/local-dev-diagnostics
Command: npx skills add https://github.com/vovanostm-public/multica --skill local-dev-diagnostics-vovanostm-public

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Local development environments fail in opaque ways: ports conflict, env files load in the wrong order, Next.js dev servers consume excessive CPU, and agent integration state drifts. This Skill replaces guesswork with deterministic diagnostic scripts that produce evidence before any code changes. ## Core Features & Use Cases - Stack Health Checks: Run dev-health.sh to verify services via /health and /health.txt endpoints, port availability, and env file loading order (.env.worktree before .env). - Agent Integration Inspection: Query task queue and agent integration database state read-only with pnpm codex:integration-inspect, optionally filtered by issue number. - Next.js and MCP Diagnostics: Profile Next.js dev CPU, memory, HMR, and missing styles, plus run MCP client/server smoke checks. - Use Case: Your worktree fails to start and styles are missing. Run the health and profiling scripts, inspect Tailwind source scanning, and identify the root cause before editing any components. ## Quick Start Run the local dev diagnostics workflow to check my Multica dev stack health and tell me why my worktree is failing to start.

Frequently Asked Questions about local-dev-diagnostics

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check local dev stack health before debugging?▼

Run bash scripts/codex/dev-health.sh to check service health, ports, and env configuration. The workflow uses /health and /health.txt endpoints rather than the root path, and loads .env.worktree before .env when both exist.

How to inspect agent integration database state read-only?▼

Run pnpm codex:integration-inspect to view task queue and agent integration DB state without modifying data. Add -- --issue-number 103 to filter by a specific issue. It uses DATABASE_URL from the environment and never logs credentials.

Why is my Next.js dev server slow or missing styles?▼

Run bash scripts/codex/profile-next-dev.sh to profile CPU, memory, and HMR behavior. If styles look missing, inspect Tailwind source scanning before editing components, and never delete .next while next-server is running.

How do I verify MCP client and server connectivity?▼

Run bash scripts/codex/mcp-smoke.sh to perform MCP client/server smoke checks. Interpret the output for evidence of connection failures before changing any MCP configuration code.

Where are Multica daemon worktrees located on disk?▼

Daemon worktrees live at ~/multica_workspaces_local/{workspace_id}/{taskKey}/workdir/multica, with the bare repository cache at ~/multica_workspaces_local/.repos/{workspace_id}/. Use these paths when diagnosing worktree startup issues.