How do I install the Preflight skill?

Run npx skillmds@latest add ryanmaclean/preflight in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Preflight safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Preflight?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Preflight free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Preflight?

ryanmaclean (@ryanmaclean) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Preflight by ryanmaclean — Agent Skill

Name: Preflight
Author: ryanmaclean

Preflight

Pre-dispatch quota check. Run before delegating tasks to Codex, Auto-Claude, or any agent backend. Shows live utilization for all providers and recommends which backend to use based on remaining quota. Warns when any provider is above 70% and blocks dispatch recommendations when above 90%.

ryanmaclean 90b8041 2.6 KB Updated Sep 17, 2026

File contents

preflight — Pre-Dispatch Quota Check

Run this before delegating batches of tasks to Codex, Auto-Claude, or any agent backend. Shows live utilization and recommends the best backend for the planned work.

Usage

/preflight
/preflight 5          # planning to dispatch 5 specs
/preflight 10 codex   # planning 10 codex tasks specifically

What it does

Reads live quota from ~/.claude/skills/providers/providers.nu --json
Evaluates each provider against thresholds (green <70%, yellow 70-90%, red >90%)
Recommends backend routing based on remaining capacity
If any critical provider is >90%, warns before proceeding

Thresholds

Level	Usage	Action
Green	<70%	Proceed freely
Yellow	70-90%	Prefer cheaper model (Haiku over Sonnet, Haiku over Codex)
Red	>90%	Do task directly or use Ollama; avoid spawning agents

Backend routing rules

Codex session >80%: Switch to Auto-Claude (Haiku) for simple tasks
Claude 5h >80%: Avoid spawning more Claude Code subagents; use Codex or Ollama
Both constrained: Do small tasks directly; defer large batches
All green: Proceed with planned dispatch

Instructions for Claude

When this skill is invoked:

Run nu ~/.claude/skills/providers/providers.nu --json via Bash tool
Parse the JSON and evaluate thresholds
Print a concise summary table with color-coded status
State clearly: which providers are constrained, which are free
Recommend: what backend to use for the planned work
If the user mentioned a task count (e.g. "/preflight 5"), estimate whether quota covers it and at what cost (Codex session burns faster than weekly)
Do NOT just print raw numbers — give an actionable recommendation

Example output

Provider         Usage    Status   Reset
──────────────────────────────────────────
Claude 5h        44%      ✓ green  2h 47m
Claude 7d        61%      ✓ green  127h
Claude 7d sonnet 70%      ⚠ yellow 129h
Codex session    21%      ✓ green  12m
Codex weekly     32%      ✓ green  95h

Recommendation: All green. Codex session resets in 12m — if dispatching >3 specs,
wait 12 minutes to start with a fresh session budget. Proceed freely otherwise.

---name: preflightdescription: Pre-dispatch quota check. Run before delegating tasks to Codex, Auto-Claude, or any agent backend. Shows live utilization for all providers and recommends which backend to use based on remaining quota. Warns when any provider is above 70% and blocks dispatch recommendations when above 90%.---# preflight — Pre-Dispatch Quota CheckRun this before delegating batches of tasks to Codex, Auto-Claude, or any agentbackend. Shows live utilization and recommends the best backend for the planned work.## Usage```/preflight/preflight 5 # planning to dispatch 5 specs/preflight 10 codex # planning 10 codex tasks specifically```## What it does1. Reads live quota from `~/.claude/skills/providers/providers.nu --json`2. Evaluates each provider against thresholds (green <70%, yellow 70-90%, red >90%)3. Recommends backend routing based on remaining capacity4. If any critical provider is >90%, warns before proceeding## Thresholds| Level | Usage | Action ||-------|-------|--------|| Green | <70% | Proceed freely || Yellow | 70-90%| Prefer cheaper model (Haiku over Sonnet, Haiku over Codex) || Red | >90% | Do task directly or use Ollama; avoid spawning agents |## Backend routing rules- **Codex session >80%**: Switch to Auto-Claude (Haiku) for simple tasks- **Claude 5h >80%**: Avoid spawning more Claude Code subagents; use Codex or Ollama- **Both constrained**: Do small tasks directly; defer large batches- **All green**: Proceed with planned dispatch## Instructions for ClaudeWhen this skill is invoked:1. Run `nu ~/.claude/skills/providers/providers.nu --json` via Bash tool2. Parse the JSON and evaluate thresholds3. Print a concise summary table with color-coded status4. State clearly: which providers are constrained, which are free5. Recommend: what backend to use for the planned work6. If the user mentioned a task count (e.g. "/preflight 5"), estimate whether quota covers it and at what cost (Codex session burns faster than weekly)7. Do NOT just print raw numbers — give an actionable recommendation## Example output```Provider Usage Status Reset──────────────────────────────────────────Claude 5h 44% ✓ green 2h 47mClaude 7d 61% ✓ green 127hClaude 7d sonnet 70% ⚠ yellow 129hCodex session 21% ✓ green 12mCodex weekly 32% ✓ green 95hRecommendation: All green. Codex session resets in 12m — if dispatching >3 specs,wait 12 minutes to start with a fresh session budget. Proceed freely otherwise.```

Frequently asked questions