Skip to content
Agentmart
Submit a tool

hunk

Open the working diff, last commit, or branch range in the Hunk terminal viewer — and with `auto`, review it inline.

Skill by feelepxyz

IndexedGPL-3.0★ 2
Write a review

Chalk Labs

Need help deploying hunk?

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/feelepxyz-hunk

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

Open the working diff, last commit, or branch range in the Hunk terminal viewer — and with `auto`, review it inline.

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
Runs as
Shell scripts

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/hunk

What's inside

Skill name
hunk
Description
Open the working diff, last commit, or branch range in the Hunk terminal viewer — and with `auto`, review it inline.
License
GPL-3.0
Instructions
56 lines in SKILL.md · view on GitHub

Before installing

  • Ships executable scripts

Files (3)

  • SKILL.md
  • openai.yaml · agents/
  • hunk.sh

SKILL.md (first lines)

# Hunk

`hunk.sh` is the whole mechanism. It resolves what is worth reviewing, opens it
in a focused Herdr tab under `--watch`, and reports the live session — one
call, a handful of lines back.

## Run it

    bash ~/.agents/skills/hunk/hunk.sh $ARGUMENTS

Run through `bash` — skills install without the exec bit.

It picks the target from the repository's own state, so pass nothing else:

| State | What opens |
| --- | --- |
| unstaged or untracked work | the working tree, untracked files included |
| staged work only | the staged changes |
| clean, on the default branch | the last commit |
| clean, on a branch | `<default>...HEAD` |
| clean, branch adds no commits | the last commit, with a `note:` saying why |

A live Hunk session for this repo is reloaded rather than duplicated. Outside
Herdr the script opens nothing and prints a `run:` line — hand that command to
the user, since the viewer is theirs.

## Follow the mode it reports

**`mode: open`** — report the `target:` and stop. The user reviews it.

**`mode: auto`** — you review it:

1. Read the Hunk skill at `$(hunk skill path)` and follow it for every session
   command below.
2. `hunk session review --repo . --json` for the file and hunk structure. Add
   `--include-patch` only for files whose diff text you actually need to read.
3. Batch every finding into one `hunk session comment apply --repo . --stdin`
   call, each anchored to the line it concerns.
4. Report the count and the single most consequential finding.

Comment where you carry information the user does not: a root cause, a broken
invariant, a case the change misses. Silence on a hunk is a valid review.

Done when every file in the `session review` listing has been read and judged,
and the batch has landed.

## When it reports an error

Every error carries a `cause:` and a `help:` line — act on the `help:` line.

## Boundary

Reviewing is the whole job. Fixing what the review finds happens when the user
asks for it.

From feelepxyz/dotfiles, GPL-3.0.

More from feelepxyz/dotfiles

Signals

Stars
2
Forks
3
Open issues
0
Last commit (this folder)
Jul 29, 2026 (3 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
GPL-3.0
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from feelepxyz

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 hunk 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 hunk?

CallGet help
Cite this page

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