Skip to content
Agentmart
Submit a tool

full-build

Implement the product in a worktree from the winning spike, one feature at a time, test-first, until the verify command passes — or, when the human says…

Skill by Kshintak

IndexedMIT
Write a review

Chalk Labs

Need help deploying full-build?

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-full-build

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

Implement the product in a worktree from the winning spike, one feature at a time, test-first, until the verify command passes — or, when the human says full-build / run the framework, resume the whole framework. Use when STATUS phase is full-build.

Use cases: testing

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/full-build

What's inside

Skill name
full-build
Description
Implement the product in a worktree from the winning spike, one feature at a time, test-first, until the verify command passes — or, when the human says full-build / run the framework, resume the whole framework. Use when STATUS phase is full-build.
License
MIT
Instructions
21 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# full-build

## If the human said "full-build" / "run the framework"

That is a request to drive the framework, not to write code now. Use `hermes-conductor` (in Hermes) or, in another tool, read `build/STATUS.md`, report the next node, and stop. Never restart a checked phase.

## As the full-build card (goal mode, worktree)

1. Setup: you are on branch `build/<idea>/full-build-c<n>`. Full tier: bring in the winning spike with `git merge --no-edit <winner branch from build/08-spikes.md>`. Quick tier (no `08-spikes.md`): start from main and build what `08-sandbox-plan.md` defines.
2. If this is cycle ≥ 2: read the top block of `build/11-verify.md` first. Those failures are your first task.
   Human decisions (`build/gates/*.approved` notes and `build/decisions.md`, all listed in your card) are features on your list too — they override PARKED assumptions and "out of scope" lines written before the human decided. Put each one on the checklist in step 3.
3. Keep a progress file at `build/runs/full-build-c<n>.md` (main checkout BUILD dir): the feature list from `09-who-what.md` (or the plan's demo definition in quick tier) with `[ ]/[x]`, commands run, decisions made. Models tamper with prose less than with JSON — keep it a checklist.
4. One feature at a time: write the failing test, make it pass, run the `verify:` command from `08-sandbox-plan.md`, commit on the worktree branch.
5. Stay inside BoB: the primary reference's approach, not the anti-pattern. Scope creep → note it in the progress file, don't build it.
6. Never delete, skip, or loosen a test to get green. The verifier checks for exactly that.
7. Only claim what you ran. Your handoff `verification` list must contain commands you actually executed with their real exit codes — a "done" that verify then fails is counted against your lane as a false claim.
8. When every feature is `[x]` and verify passes locally: request review (the card body names a different-vendor reviewer). Put the verify command and exit code in your handoff metadata.

Builders never push, merge to main, deploy, or touch credentials. Blocked on a product question → `kanban_block` with the question.

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 full-build 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 full-build?

CallGet help
Cite this page

full-build skill. Agent Mart. https://agentmart.dev/skills/kshintak-full-build (last updated 2026-10-24)