Skip to main content
For teams already on Claude Code
·Built for Claude Code

A folder we set up for you. Real work done.

You already pay for Claude. These kits let it earn its keep. We set up the folder inside Claude Code, run the first-run interview, configure the tools it needs, and forge custom skills tuned to your business. No dashboards. No logins. No SaaS to renew. You own what we build.

WE SET IT UP FOR YOU·FROM $500·RUNS ON YOUR MACHINE

How a kit gets set up.

We set up the folder and the process for you, then hand it over. You own everything we build, with no subscription to renew.

We set it up for you

From $500

Discovery call first, scoped quote after. The base $500 covers a single kit, standard MCPs, and a baseline forge pass. We quote anything beyond that up front.

What's included
  • 30-min discovery call to pick the right kit and clarify the use case
  • Kit installation plus standard MCP and credential setup
  • First-run interview run together or async
  • One forge pass tuned to your business
  • 30-min training and walkthrough call
  • 2 weeks of email support after handoff
Quoted on top (after discovery)
  • +Custom skills beyond the baseline forge
  • +Extra MCPs or third-party integrations
  • +Helper scripts or Python additions
  • +Scheduling setup (cron, GitHub Actions, and similar)
  • +Multiple kits bundled together

Available kits

6 KITS
MVP

website-audit

Crawls your site, finds orphan pages, surfaces hub pages, renders an interactive link graph.

Point it at a domain. It crawls via Tavily, builds a link graph in JSONL + CSV, finds orphan pages and hub pages, and renders a self-contained interactive HTML viz you can open in any browser. Schedulable.

RUN   /audit
FOR   SEO leads, content strategists, site owners
REQ   TAVILY_API_KEY
MVP

competitor-monitor

Snapshots competitors weekly. Diffs against last run. Writes a narrative digest.

Give it a list of competitors. Each week it snapshots their pricing, changelogs, blogs, careers, and leadership pages, diffs against the last snapshot, and writes a narrative digest of what changed and why it matters. SQLite under the hood. Schedulable.

RUN   /digest
FOR   Founders, product leads, sales, strategy
REQ   WebFetch + Chrome MCP fallback
MVP

research-kit

Multi-agent research engine. Decomposes a question, runs parallel researchers, builds a knowledge graph.

Ask a research question. The kit decomposes it, dispatches up to 5 parallel researcher subagents, persists findings into a per-run knowledge graph (sources, claims, themes, entities), and produces a multi-format report with full source traceability and an interactive KG viz.

RUN   /research <question>
FOR   Strategy teams, consultants, analysts
REQ   pandoc on PATH (for docx/pdf)
MVP

lead-gen

Scores accounts against your ICP, researches the strongest, drafts personalized 3-touch outreach.

Drop in a CSV of candidate accounts. The kit scores them against your ICP, researches the top scorers for triggers (funding, hires, jobs, launches, press), and drafts personalized 3-touch outreach in your voice. Never sends, drafts only. Schedulable.

RUN   /scout then /draft <account>
FOR   Founders, SDRs, account executives
REQ   WebFetch + Chrome MCP fallback
New

site-pulse

Weekly perf digest from Cloudflare, GA4, Search Console, PageSpeed. Auto-drafts GitHub issues.

Pulls metrics from Cloudflare, GA4, Search Console, and PageSpeed Insights, runs AIO checks, and writes a weekly digest of what changed. Triage step turns the most important findings into draft GitHub issues you can review and ship.

RUN   /pulse then /triage
FOR   Site owners, growth engineers, product teams
REQ   Cloudflare / GA4 / GSC / PSI credentials per source
New

doc-intake

Local-first document extraction with an editable UI. Your documents never leave your machine.

Drop in PDFs, scans, spreadsheets, contracts. The kit extracts structured data, hands you an editable review UI to verify, and exports clean tables to CSV/JSON. All processing local, no documents leave your machine.

RUN   /intake
FOR   Ops, finance, legal, anyone re-keying PDFs
REQ   Local Python 3, no external API needed

How we set it up.

01

Discovery call

We start with a short call to pick the right kit, clarify the outcome you want, and scope anything beyond the base setup before we begin.

02

We install and configure it

We set up the kit folder inside Claude Code on your machine, run the first-run interview, and configure the MCPs and credentials it needs. No plugin to register, the folder is self-contained.

03

We forge skills tuned to you

We crystallize what the kit learns about your business into custom skills, written into your copy. The kit gets sharper for your work over time, and you own it.

04

Training and handoff

We walk your team through running the kit on demand, set up scheduling if you want it, and hand everything over with two weeks of email support. It is yours to keep.

When a kit isn't enough.

Kits run on your machine. The moment you need a hosted dashboard, multiple users, real-time integrations, or a custom UI, that's where a pre-built app or custom build comes in. We'll tell you when you've outgrown a kit.