Skip to content
Agentmart
Submit a tool

box-it-up

Plan, build, and compare isolated spikes — the smallest falsifiable demo, built in parallel git worktrees by different vendors.

Skill by Kshintak

IndexedMIT
Write a review

Chalk Labs

Need help deploying box-it-up?

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/kshintak-box-it-up

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

Plan, build, and compare isolated spikes — the smallest falsifiable demo, built in parallel git worktrees by different vendors. Use when STATUS phase is box-it-up or the human says sandbox, worktree, spike, or box it up.

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/box-it-up

What's inside

Skill name
box-it-up
Description
Plan, build, and compare isolated spikes — the smallest falsifiable demo, built in parallel git worktrees by different vendors. Use when STATUS phase is box-it-up or the human says sandbox, worktree, spike, or box it up.
License
MIT
Instructions
31 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# box-it-up

Three steps; your card's `Args.step` says which one you are.

## step: plan

Read BoB + assumptions + human decisions (`build/gates/*.approved` notes and `build/decisions.md` — binding, they override PARKED items). **Quick tier** (`Args.tier: quick`): there is no BoB or spike — plan straight from `00-intake.md` + `01-assumptions.md`, and make the demo definition the whole deliverable. Write `build/08-sandbox-plan.md` (`~/.build-framework/templates/build/08-sandbox-plan.md`):

- Demo definition — the smallest thing that would falsify the idea
- `verify: <one shell command>` — exit 0 means the demo works (this line is required)
- Out of scope (explicit)
- Budget per spike (default `router.yaml budgets.spike_max_usd_each`)

## step: spike (one card per spike, parallel, each in its own worktree)

1. You are in an isolated worktree on branch `build/<idea>/spike-<name>-c<n>` (card body). Never touch main.
2. Build only the demo. No extra features, no refactors outside the demo.
3. Run the `verify:` command. If red, patch up to 3 times, then stop.
4. Commit on the worktree branch (local only; never push).
5. Write `build/spikes/<name>.md` in the **main checkout's** BUILD dir: result (PASS/FAIL), verify command + exit code, `git diff --stat` vs base, branch, worktree path, cost, what you would do next.
6. Request review as the card body says (a different vendor reviews). This is a goal-mode card: the judge keeps you going until the report and verify evidence exist.

## step: compare

1. Read `build/spikes/*.md` and each branch's diff (`git diff <base>...<branch> --stat`).
2. Pick a winner against the plan's demo definition — not the prettier code. A spike whose verify failed cannot win unless every spike failed (then say so).
3. Write `build/08-spikes.md`: winner, **branch**, why, and the loser's one idea worth keeping. (Hermes removes spike worktrees after this card; the branches stay.)
4. End with the human instruction for the gate:
   > `git -C <repo> worktree add .worktrees/review-<winner> <winner branch>`, open that folder in Cursor, and review the diff against `build/08-sandbox-plan.md` (the `review-diffs` rule helps). Then `hbs approve gate-cursor --note "..."`.

From Kshintak/new-build-framework, MIT.

More from Kshintak/new-build-framework

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
Sep 29, 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 Kshintak

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 box-it-up 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 box-it-up?

CallGet help
Cite this page

box-it-up skill. Agent Mart. https://agentmart.dev/skills/kshintak-box-it-up (last updated 2026-10-24)