Skip to content
Agentmart
Submit a tool

implement

Build one issue through three fresh-eyes phases — build, harden, gaps — write its docs with you, and report only what needs your eye; stops with the tree dirty.

Skill by javierponferradalopez

IndexedMIT
Write a review

Chalk Labs

Need help deploying implement?

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/javierponferradalopez-implement

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

Build one issue through three fresh-eyes phases — build, harden, gaps — write its docs with you, and report only what needs your eye; stops with the tree dirty.

Use cases: documents, 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. Checked Oct 25, 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/implement

What's inside

Skill name
implement
Description
Build one issue through three fresh-eyes phases — build, harden, gaps — write its docs with you, and report only what needs your eye; stops with the tree dirty.
License
MIT
Instructions
97 lines in SKILL.md · view on GitHub

Files (4)

  • BUILD.md
  • HARDEN.md
  • SKILL.md
  • STANDARDS.md

SKILL.md (first lines)

# Implement

You are the **orchestrator**. You build **one** issue with three skills — `/build`,
`/harden`, `/gaps` — one after the other. Each one runs in its own sub-agent, so it
has **fresh eyes**: it never sees the thinking of the step before it. Your window
holds short returns only.

**Argument:** an issue reference — `/implement #N`, a URL, or a path, depending on the
repo's tracker — plus optional **extra instructions**, such as `completa los huecos 1
y 3`. Numbers refer to the last report of this session: resolve each one to the
requirement text it names.

## 1. Run the steps

Send the user one short line as each step starts.

1. **Build** — a foreground `general-purpose` sub-agent. Brief: "Invoke `/build #N`
   `<extra instructions>`. You run under `/implement`." Its return starts with one of:
   - **blocked** → put its question to the user in prose, and send the answer to the
     **same** sub-agent (`SendMessage`), so it keeps the context it loaded.
   - **red** → stop and report what fails. Hardening red code hides the cause.
   - **done** → keep its **decisions** for the report.
2. **Harden** — a foreground `general-purpose` sub-agent. Brief: "Invoke `/harden #N`
   on the dirty tree. You run under `/implement`." Skip it when the diff holds no
   code — only docs, ADRs, or config with no logic.
3. **Gaps** — a foreground `general-purpose` sub-agent. Brief: "Invoke `/gaps #N` on
   the dirty tree. You run under `/implement`." It judges the final diff, so it runs
   last and runs even when harden ended red.

The orchestrator never fills a gap: a requirement that is missing stays missing until
the user names it in the extra instructions of a later run.

## 2. Report — filter for attention

End with the report the user reads, in their configured language. It is a filter,
not a log: it holds only what the user must decide or look at.

- **One status line**: typecheck / lint / test (green or red), plus counts — tests
  harden added, deferred requirements.
- **Look closer** — only when there is something, one line each with `file:line`:
  - each **decision** build made, with its reason;
  - each **bug harden fixed**: a behavior change nobody asked for;
  - each **real gap** — an owned requirement that is missing, partial or wrong —
    numbered, so the user can name it in the next run;
  - **unasked** behavior, **unowned** spec requirements and **ambiguous** ones;
  - what is still **red**.

An empty **Look closer** leaves the status line as the whole report. What was built
and the deferred list stay out of it.

## What this does not do

Commit, push, create a branch, or close the issue. Work on the current branch and
leave the tree dirty.

From javierponferradalopez/skills, MIT.

More from javierponferradalopez/skills

Signals

Stars
0
Forks
0
Open issues
1
Last commit (this folder)
Oct 23, 2026 (7 days ago)

From GitHub (fetched Oct 24, 2026).

Details

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

More from javierponferradalopez

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 implement 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 implement?

CallGet help
Cite this page

implement skill. Agent Mart. https://agentmart.dev/skills/javierponferradalopez-implement (last updated 2026-10-24)