Skills Explorer Add Skills

vercel-optimize

An agent skill by vercel-labs, from vercel-labs/agent-skills. Tags: debugging, devops, performance, trend-analysis, web.

What it does

Use for Vercel cost and performance optimization on deployed projects, especially Next.js, SvelteKit, Nuxt, and limited Astro apps. Collect Vercel metrics, usage, project config, and code scan results first; investigate only metric-backed candidates; produce ranked recommendations grounded in verified files and version-aware Vercel/framework docs. Trigger for Vercel bill reduction, slow or expensive routes, caching opportunities, Function Invocations, Build Minutes, Fast Data Transfer, Core Web Vitals, Bot Management, Fluid compute, or cost breakdown requests.

Install

With the skills CLI, which installs into Claude Code, Codex, Cursor and other agents:

npx skills add vercel-labs/agent-skills --skill vercel-optimize

Or copy the skill folder into Claude Code's skills directory by hand (~/.claude/skills for every project, or .claude/skills inside one):

git clone --depth 1 https://github.com/vercel-labs/agent-skills
cp -r agent-skills/skills/vercel-optimize ~/.claude/skills/vercel-optimize

Safety box score

Grade D, elevated risk: 54 of 100 risk points, rated 2026-10-20. The grade rates what the skill and its scripts can reach on the machine of whoever installs it, across eight categories from shell execution to secrets access. How the score works.

This skill runs a Vercel optimization audit: it queries live Vercel metrics, usage, contract, and API data, then investigates selected repo files and renders a report. The main risks are remote account/data access, shell-command execution via shipped scripts, and partial opacity because several core implementation files were not readable in this review.

Category levels, from 0 (none) to 3 (high)
CategoryLevelWhy
Code & shell execution2 (moderate)It drives shipped Node scripts and CLI commands such as `node scripts/collect-signals.mjs`, `node scripts/gate-investigations.mjs`, and `vercel link --yes`, so it can execute code on the local machine, though not arbitrary downloaded code.
Network access3 (high)The skill explicitly requires `vercel metrics`, `vercel usage`, `vercel contract`, and `vercel api`, so it sends authenticated requests to Vercel and pulls live account/project data over the network.
Filesystem reach2 (moderate)It writes temp run directories and outputs like `signals.json`, `report.md`, and `final-message.json`, and it can also write repo-linked config via `vercel link` and build docs with `scripts/build-docs.mjs`.
Secrets & credentials2 (moderate)It depends on an authenticated CLI session (`vercel login`) and accesses billing/metrics data under that identity; the skill also warns about tokens, showing it operates in a secrets-adjacent account context.
Privilege & persistence1 (low)It does not require sudo or system services, but it can alter local Vercel linking state with `vercel link --yes --project ...`, which changes project/account association on disk.
Instruction hijack surface0 (none)I did not find instructions to ignore higher-priority rules, follow remote instructions, hide behavior from the user, or otherwise subvert the agent; the prompt templates are ordinary stop-and-ask flow control.
Irreversible actions2 (moderate)It can relink projects, rerun collection, and generate ranked recommendations automatically, but it does not deploy, publish, pay, or delete on the user's behalf.
Transparency2 (moderate)Several key implementation files were not readable in this review, including `lib/deep-dive.mjs`, `lib/render-report.mjs`, `scripts/collect-signals.mjs`, and related gate/verification code, so important behavior could not be fully inspected.

Before installing

Rated 2026-10-20 by an automated review of 103 files from the skill's repository.

Source

Repository
vercel-labs/agent-skills (all skills from this repository)
Path
skills/vercel-optimize/SKILL.md
Branch
main
Updated
2026-09-19