Agent Starter Kit
7 battle-tested agent skills + persona templates. Free.
7 skills · 28 failure modes
View spec
Get free →WORKS WITH · AGENT SKILLS STANDARD
SKILL.md is an open standard adopted by 30+ agents by mid-2026 — GitHub Copilot, Cursor, Windsurf, Gemini CLI, Goose, OpenCode and more. If your tool reads SKILL.md, our kits work.
WHERE SKILLS LIVE
Wherever your tool loads skills from — consult its docs for the skills directory; the SKILL.md files themselves need no changesHOW THEY TRIGGER
Per your tool — most match on the skill's description automatically.
RUNNING IT UNATTENDED
Whatever your tool provides, or cron around its headless mode
The skills standard covers what an agent knows, not when it wakes up — scheduling is each tool's own concern. If yours has a headless mode, the same wrapper pattern our kits use for codex exec applies unchanged.
WITH YOUR LAPTOP CLOSED
Only with a host that stays up
PLATFORM DOCS: agentskills.io
7 battle-tested agent skills + persona templates. Free.
7 skills · 28 failure modes
View spec
Get free →Field manual: 150 production prompts with failure-mode annotations for each.
View spec
43 production agent skills + the framework to write your own.
43 skills · 219 failure modes
View spec
Cold outreach that runs itself: enrichment → sequence → reply triage.
2 skills · 9 failure modes
Runs unattended
View spec
Agents that run your site after it's built — and tell you when they didn't.
9 skills · 3 templatesneeds an API key
Runs unattended
View spec
Rank on Google. Get cited by ChatGPT. Same pipeline.
13 skills · 64 failure modes · 3 templates
Runs unattended
View spec
Every account runs like your best account — onboarding, reports, invoices, scope, health, offboarding.
7 skills · 36 failure modes · 6 templates
Runs unattended
View spec
ZYTE
Scraping that notices when it silently failed — JS rendering, pagination, and the 200-with-zero-rows trap.
1 skill · 7 failure modesneeds an API key
View spec
XATA
Database work your agent won't get wrong — schema design, full-text search, and the queries that quietly return nothing.
1 skill · 7 failure modesneeds an API key
View spec
WACHETE
Change detection that doesn't cry wolf — real diffs, not layout noise. Works with no API key.
1 skill · 6 failure modesno API key needed
View spec
ZEROBOUNCE
Clean the list before you burn the domain — validation, spam traps, and what a risky score actually means.
1 skill · 5 failure modesneeds an API key
View spec
WHOISFREAKS
Domain intelligence with its dates intact — WHOIS, ownership history, expiry, and privacy-masked records read correctly.
1 skill · 3 failure modesneeds an API key
View spec
VIRUSTOTAL
Check before you download — 70+ engines, and how to read a result when three of them disagree.
1 skill · 5 failure modesneeds an API key
View spec
WHOP
Sell without mispricing — product creation, checkout links, and price as a stop-and-confirm field.
1 skill · 4 failure modesneeds an API key
View spec
ZYLVIE
Cross-list without overwriting — orders, listings, and why a blind write costs you the fields you never read.
1 skill · 5 failure modesneeds an API key
View spec
WIX
Build Wix sites programmatically — pages, components, and publishing as the deliberate, irreversible step.
1 skill · 4 failure modesneeds an API key
View spec
WATI
WhatsApp that won't double-send — templates, contacts, and why retrying after a timeout is the dangerous move.
1 skill · 4 failure modesneeds an API key
View spec
ZIXFLOW
CRM your agent keeps clean — contacts, deals, pipelines, and the duplicate records nobody notices creating.
1 skill · 4 failure modesneeds an API key
View spec
WOLFRAM ALPHA
Computation with units carried through — math, statistics, and reading the interpretation before trusting the answer.
1 skill · 5 failure modesneeds an API key
View spec
WORLD NEWS
News with its dates and sources attached — headlines, filters, and why one outlet is never consensus.
1 skill · 5 failure modesneeds an API key
View spec
WRITER
Brand voice without invented facts — tone control, and checking the claims rather than the prose.
1 skill · 4 failure modesneeds an API key
View spec
WIT.AI
Intent detection that respects confidence — entities, thresholds, and never acting decisively on a 0.3 match.
1 skill · 4 failure modesneeds an API key
View spec
WEBSCRAPING.AI
Simple scraping, honestly reported — extraction, proxies, and asserting on row count instead of status code.
1 skill · 5 failure modesneeds an API key
View spec
YELP
Business data your agent won't misquote — ratings carried with their sample size, locations resolved before searching.
1 skill · 5 failure modesneeds an API key
View spec
ZEP
Memory that doesn't leak between users — persistent facts, session scoping, and pruning before retrieval degrades.
1 skill · 4 failure modesneeds an API key
View spec
WINSTON
AI detection read responsibly — scores with thresholds, length limits, and why one run is never a verdict.
1 skill · 4 failure modesneeds an API key
View spec
ZULIP
Team chat your agent won't spam — streams, topics, and treating a send as the irreversible act it is.
1 skill · 5 failure modesneeds an API key
View spec
ZEPLIN
Design handoff without invented values — tokens over pixels, versions pinned, gaps reported instead of guessed.
1 skill · 4 failure modesneeds an API key
View spec
Closed-loop growth for any site — seven sources, and the one that silently returned nothing.
1 skill · 5 failure modes · 2 templatesneeds an API key
View spec
Nine controls that refuse the commands that deleted other people's work.
View spec
Scrape → Store → Monitor. Zyte + Xata + Wachete in one bundle. Save $13.
needs an API key
View spec
Verify → Reach → Track. ZeroBounce + Wati + Zixflow in one bundle. Save $18.
needs an API key
View spec
List → Sell → Build. Whop + Zylvie + Wix MCP in one bundle. Save $18.
needs an API key
View spec
All 5 workflow kits + 12 months of new workflow kits. One price.
View spec
A $1 end-to-end test of the purchase chain. Not for sale to the public.
1 skill · 5 failure modes
View spec
VERIFIED = installed and exercised by us on this platform · COMPATIBLE = via the open SKILL.md standard
TRY BEFORE YOU SPEND A CENT
The free Agent Starter Kit installs on 30+ tools via the open standard in minutes — 7 production skills, real setup docs.
OTHER PLATFORMS: Claude Code/Codex CLI/Hermes Agent