build
Implement one ticket end-to-end and leave the working tree dirty: no commit, push, branch, or ticket close.
Skill by noel-lopez
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/noel-lopez-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.
About
Reviews
Write a reviewNo 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 Oct 26, 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/build
What's inside
- Skill name
build- Description
- Implement one ticket end-to-end and leave the working tree dirty: no commit, push, branch, or ticket close. Use for `/build #N` (or a ticket URL/path).
- License
- MIT
- Instructions
- 69 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Build Build **one** ticket end-to-end, then **stop and leave the working tree dirty**. This is one deliberate step in a human-in-the-loop flow. **Argument:** a ticket reference — `/build #N`, a URL, or a path, depending on the repo's tracker. Implement **only** that ticket. ## 1. Fetch the ticket The issue tracker convention should already be in your context. If it isn't, stop and report that `/setup-matt-pocock-skills` needs to run first. Fetch the given ticket **with its full body and comments**, and pull in its parent spec if it has one. ## 2. Load context Before writing code, read the project's own bar and shape: - `GLOSSARY.md`, `docs/adr/`, `CLAUDE.md` / `AGENTS.md`. - Explore the repo and fill context with the parts relevant to this ticket — **especially the test files** that touch the area you'll change. ## 3. Apply the quality bar Invoke **`/coding-standards`** for the bar on deep modules, testing, and clean code. Don't restate standards here — that skill owns them. ## 4. Implement (TDD where feasible) Derive the behaviors to build from the ticket's **acceptance criteria**. Apply red-green-refactor where it's feasible — one cycle at a time: 1. **RED** — write one failing test for the next behavior. 2. **GREEN** — implement just enough to pass it. Never refactor while RED. 3. **REPEAT** — next test responds to what the last cycle taught you. Don't write all the tests up front. 4. **REFACTOR** — once green, clean up against the coding-standards bar. Embody this discipline directly. Where TDD doesn't fit, just build it well; a later review pass is the safety net that adds tests. ## 5. Autonomy & escalation Work autonomously — no plan-approval gate. Escalate to whoever invoked you (the user or an orchestrating agent), **in prose, no interactive prompts**, **only** for a blocking ambiguity that neither the ticket, its spec, `GLOSSARY.md`, nor the ADRs resolve. Don't guess blindly; don't ask about everything either. As a subagent, escalating means stopping and putting the open question in your final summary. ## 6. Feedback loop Before stopping, run the repo's **real** typecheck and test commands and get them green. Get them from where the repo documents them — `CLAUDE.md` / `AGENTS.md`, project memory, a `docs/` runbook. Only if they aren't written down anywhere, fall back to discovering them (`package.json` scripts, `Makefile`, README); don't assume `npm`, and don't spelunk in a loop. ## 7. Stop — leave it dirty When the ticket is done: - **Do NOT** commit, push, create a branch, or close the ticket. Work on the current branch; branches are the user's concern. - End with a short summary: what you built, key decisions, and the typecheck/test status.
From noel-lopez/skills, MIT.
More from noel-lopez/skills
- 0.609score
coding-standards
noel-lopez
The universal bar for good code — deep modules, design for testability, behavior-driven tests, mocking only at system boundaries, lean clean code.
SkillAFreshness: A★ 3Claude Code +9 - 0.601score
improve
noel-lopez
Review one ticket's uncommitted implementation in a fresh context: fix and harden in place, flag spec gaps for the human, never commit.
SkillAFreshness: A★ 3Claude Code +9 - 0.566score
walkthrough
noel-lopez
Build a throwaway HTML reading guide for a set of changes (dirty tree, local commits, or a PR): what it means for users and for the system, then an ordered…
SkillBFreshness: B★ 3Claude Code +9 - 0.561score
prune-comments
noel-lopez
Prune the comments in freshly implemented code — every comment earns its place under one test, or it goes.
SkillBFreshness: B★ 3Claude Code +9 - 0.558score
resolve-conflicts
noel-lopez
Resolve conflict markers left by a rebase, merge or cherry-pick, and orchestrate a stacked rebase.
SkillBFreshness: B★ 3Claude Code +9 - 0.519score
check-upstream
noel-lopez
Audits the manifest's tracked copies of Matt Pocock's skills against his upstream repo in two separate passes — main (his released state) and the open PRs.
SkillCFreshness: C★ 3Claude Code +9 - 0.514score
commit
noel-lopez
Split a dirty working tree into an ordered list of atomic conventional commits, then commit them only on an explicit literal OK.
SkillCFreshness: C★ 3Claude Code +9
Signals
- Stars
- 3
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 15, 2026 (14 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/noel-lopez/skills
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from noel-lopez
- 0.609score
coding-standards
noel-lopez
The universal bar for good code — deep modules, design for testability, behavior-driven tests, mocking only at system boundaries, lean clean code.
SkillAFreshness: A★ 3Claude Code +9 - 0.601score
improve
noel-lopez
Review one ticket's uncommitted implementation in a fresh context: fix and harden in place, flag spec gaps for the human, never commit.
SkillAFreshness: A★ 3Claude Code +9 - 0.566score
walkthrough
noel-lopez
Build a throwaway HTML reading guide for a set of changes (dirty tree, local commits, or a PR): what it means for users and for the system, then an ordered…
SkillBFreshness: B★ 3Claude Code +9 - 0.561score
prune-comments
noel-lopez
Prune the comments in freshly implemented code — every comment earns its place under one test, or it goes.
SkillBFreshness: B★ 3Claude Code +9 - 0.558score
resolve-conflicts
noel-lopez
Resolve conflict markers left by a rebase, merge or cherry-pick, and orchestrate a stacked rebase.
SkillBFreshness: B★ 3Claude Code +9 - 0.519score
check-upstream
noel-lopez
Audits the manifest's tracked copies of Matt Pocock's skills against his upstream repo in two separate passes — main (his released state) and the open PRs.
SkillCFreshness: C★ 3Claude Code +9
Cite this page
build skill. Agent Mart. https://agentmart.dev/skills/noel-lopez-build (last updated 2026-10-24)