Skip to content
Agentmart
Submit a tool

mc-scan

Onboard madcompany onto an existing codebase (brownfield mode).

Skill by lokkrish

IndexedMIT★ 1
Write a review

Chalk Labs

Need help deploying mc-scan?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/lokkrish-mc-scan

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Onboard madcompany onto an existing codebase (brownfield mode). Scan the stack, commands, modules, conventions and third-party services, write a codebase map the team reads before working, set the real checks, and file Human help for missing keys. Use once when adding madcompany to an app that already exists.

Use cases: writing

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md and code. Checked Nov 1, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/mc-scan

What's inside

Skill name
mc-scan
Description
Onboard madcompany onto an existing codebase (brownfield mode). Scan the stack, commands, modules, conventions and third-party services, write a codebase map the team reads before working, set the real checks, and file Human help for missing keys. Use once when adding madcompany to an app that already exists.
License
MIT
Instructions
21 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Scan an existing codebase

You act as the madcompany **lead** (`as: "lead"`). HQ must be running.

1. **Quick scan.** Run `npx madcompany scan --json`. It reports languages, package manager, frameworks, workspaces, folders by size, the check commands it found, third-party services from the dependencies, env keys from `.env.example`, CI workflows and docs. It never reads `.env`.
2. **Read the code that matters**: the README and any CLAUDE.md/AGENTS.md, the entry points, routing, data layer, one or two representative modules per area, the tests, and the CI workflow. If BMad is installed, `/bmad-project-context` can help; don't repeat what it already wrote.
3. **Write the codebase map** with `mc_design_write` at `docs/design/codebase/overview.md`, short and linked:
   - stack and how to run it locally (commands, ports, env keys by name)
   - modules and folders, one line each, with the owner domain for each (this becomes the team's domains)
   - conventions to follow: naming, state management, API style, error handling, test style, lint rules
   - risky areas and known debt: things not to touch without asking
   - a Mermaid diagram of the main flows if it helps
4. **Requirements.** There's no PRD, so record what the app already does: one `mc_requirement` per user-facing capability you found (kind `FR`, `source` the codebase map), so every later change can say what it amends or extends.
5. **Checks.** Put the real typecheck, lint and test commands in `checks:` in `.madcompany/team.yaml`, and run them once. If they fail on the current code, say so in one line and file it as the first ticket; the team can't gate on checks that are already red.
6. **Tools.** HQ → Tools suggests MCP servers for what the scan found (e.g. Sentry or Supabase if the code uses them). Mention the useful ones to the human; each is a Human help item they can add.
7. **Human help.** For each service the code uses whose keys aren't in place (the human can see which in HQ), call `mc_human_help` with steps and the key names. Same for access you need: staging URLs, test accounts, a copy of real data, CI secrets.
8. **Team.** Tell the human which domains the map suggests (e.g. "web: app/ and components/; API: server/; data: prisma/") and hand over to `/mc-staff`.
9. **First release.** Then `/mc-plan-release`: R1 is onboarding plus one small real change, so the team proves it can ship here before bigger work.

Rule for everyone afterwards: follow the existing conventions, keep changes in scope, and never rewrite working code that isn't part of the ticket.

From lokkrish/madcompany, MIT.

Signals

Stars
1
Forks
0
Open issues
0
Last commit (this folder)
Sep 25, 2026 (1 month ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from lokkrish

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get mc-scan running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying mc-scan?

CallGet help
Cite this page

mc-scan skill. Agent Mart. https://agentmart.dev/skills/lokkrish-mc-scan (last updated 2026-10-24)