Vercel Eve Agent Skills - Setup Guide¶

Source: skills.sh (4 skills, ~6.9K combined installs) GitHub: vercel/eve (5,042⭐, 527 forks, Apache-2.0, pushed Sep 11, 2026 - same-day active) Category: Agent Development / Backend First Seen: Jun 17, 2026 on skills.sh (surfaced in the Sep 11, 2026 sweep) Quality Tier: 🟡 Trusted (Gen Agent Trust Hub Pass / Socket Warn / Snyk Warn on the flagship skill)

Four skills published from Eve - Vercel's open framework for building durable backend AI agents. The flagship eve skill teaches an agent to build, edit, and debug Eve projects (agent instructions, skills, tools); the other three are the framework's own engineering-hygiene skills for PR descriptions, technical writing, and human-sounding prose. Drafted below the 20K install bar on first-party platform-org authority (the vercel org, 5K-star repo, same-day commits).

Installation¶

npx skills add vercel/eve

Skill file locations in the repo: skills/eve/SKILL.md and .agents/skills/{gh-pr-description,technical-writing}/SKILL.md. Copy any of them into an agent skills directory to use without the CLI:

git clone https://github.com/vercel/eve.git
cp eve/skills/eve/SKILL.md ~/.hermes/skills/eve/

Skill Roster¶

Skill	Installs	What It Does
`eve`	5.5K	Build durable backend AI agents with the eve framework - agent instructions, skills, tools, and debugging; points at `node_modules/eve/docs/` reading order before writing code
`gh-pr-description`	661	Generate structured GitHub PR descriptions
`technical-writing`	623	Documentation quality with content-type, editing, style, and review references
`human-writing`	145	Writing that does not read as AI-generated

Quick Start¶

Install the suite - no API keys required
eve skills any eve project: create agents, wire skills and tools, debug; the skill insists on reading the framework docs first
gh-pr-description and technical-writing work standalone on any repo or docs task

CorpusIQ Use Cases¶

Use Case	How
Backend agent stack	Eve's durable-agent architecture is a direct reference for CorpusIQ's own agent infrastructure decisions
PR and docs hygiene	`gh-pr-description` and `technical-writing` are immediately useful on the corpusiq-docs and MCP-server repos
Content voice	`human-writing` aligns with CorpusIQ's anti-AI-slop content standards

Integration with CorpusIQ¶

CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.

This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.

See what CorpusIQ connects

Limitations / Verification¶

Sub-20K bar, platform authority: 6.9K installs is below the 20K drafting bar; drafted on first-party platform-org authority (official vercel org - the Aug 16 precedent for vendor orgs) plus 5,042⭐ and same-day repo activity.
Two audit Warnings: the flagship eve skill renders Socket Warn and Snyk Warn - review-first for anything beyond the docs-hygiene skills.
Framework coupling: the eve skill is only useful on eve projects; the three hygiene skills are the general-purpose value.

# Verify skill installed
ls ~/.hermes/skills/eve/SKILL.md

Security¶

Audit	Verdict
Gen Agent Trust Hub	Pass
Socket	Warn
Snyk	Warn

← Skills Catalog | Marketplace → Powered by CorpusIQ