clean-comments
Rewrite the existing code comments in a repo or path so each says only what the code is for, and list what needs the user's judgement (comments that disagreed…
Skill by GUOJIE-HONG
Chalk Labs
Need help deploying clean-comments?
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.
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/guojie-hong-clean-comments
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
Use cases: documents, frontend, productivity
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. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
- Runs as
- Node.js package
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/clean-comments
What's inside
- Skill name
clean-comments- Description
- Rewrite the existing code comments in a repo or path so each says only what the code is for, and list what needs the user's judgement (comments that disagreed with the code, unclear names, TODO/FIXME, commented-out code) in an HTML report.
- License
- MIT
- Instructions
- 81 lines in SKILL.md · view on GitHub
Before installing
- Ships executable scripts
Files (11)
- SKILL.md
- openai.yaml · agents/
- fira-sans-400.woff2 · assets/
- fira-sans-600.woff2 · assets/
- fira-sans-700.woff2 · assets/
- fira-sans-OFL.txt · assets/
- report.css · assets/
- build-report.mjs · scripts/
- LICENSE · vendor/highlight.js/
- UPSTREAM.txt · vendor/highlight.js/
- highlight.min.mjs · vendor/highlight.js/
SKILL.md (first lines)
# Clean Comments
The user treats the code and its comments as the project's spec, and so do the agents that work on it. A comment that sends the reader elsewhere, or says what the code does not do, misleads both. You rewrite the comments that already exist, in place, and hand the user a report of what only they can decide.
If the user passed a path, work only under it; otherwise the whole repo.
## Before you start
If `git status` shows uncommitted changes, stop and ask the user to commit or stash them, so your rewrites land in a diff of their own. Never commit when you finish.
## Which comments to rewrite
Rewrite a comment that:
- carries an **Outside reference**: it points the reader to a document, spec section, ticket, meeting, discussion, or date instead of saying what the code is for;
- says something the code does not do;
- records history or discussion;
- refers vaguely to other code ("same as the service above"): name the class.
Keep a doc comment that only restates the name (`/// <summary>日期格式。</summary>`, `/** Date format. */`), and keep pointers to other code such as `<see cref="..."/>` or `{@link Mailer}`.
Do not open the documents a comment points to. Write the new comment from the code alone: why it exists and what its flow is, as short as it can be (`SMTP寄信通用服務`, `查詢BPM表的審核狀態`), in the comment's original language. When nothing is worth saying, delete the comment.
Change comments only: no code, no names, and no new comments where there were none.
Skip generated and third-party files: `Migrations/`, `*.Designer.cs`, `*.pb.go`, `node_modules/`, `vendor/`, and files that open with a license.
## Findings
Four things are the user's call. Record each in the findings file; an ordinary rewrite is not recorded, since `git diff` shows it.
- `mismatch`: a comment disagreed with the code. Rewrite it to match the code and keep the old comment's exact text as `oldComment`; the code may be what is wrong.
- `naming`: a method, variable, or class whose name does not say what it is for. Leave the name; the note says why it is unclear.
- `todo`: a `TODO` or `FIXME`. Leave it as is.
- `commented-out`: code that was commented out. Leave it as is.
Write the findings file to the system temp directory with a file-writing tool, after you finish editing, so its line numbers match the files as you leave them:
```json
{
"repo": "ErpPortal",
"scope": "src/",
"lang": "zh-TW",
"findings": [
{
"kind": "mismatch",
"file": "src/Mail/SmtpMailService.cs",
"lines": [45, 45],
"show": [[40, 47]],
"note": "舊註解說失敗會重試 3 次,程式只寄一次就拋出例外。",
"oldComment": "// 寄信失敗時自動重試三次,每次間隔 5 秒"
},
{
"kind": "naming",
"file": "src/Bpm/Helper.cs",
"lines": [9, 9],
"show": [[8, 10], [14, 14], [19, 19], [32, 32], [44, 44]],
"note": "Helper 看不出負責什麼;它只查詢 BPM 表的審核狀態。"
}
]
}
```
- `scope` is the path the user gave, or `.` for the whole repo.
- Ranges include both ends. `lines` are the lines the finding is about; the report marks them. `show` is what the row shows when opened: the whole method or property as one range, or, for a long class or interface, its declaration plus each member's signature.
- `note` is the one sentence the user reads before opening the row, written in the report's language.
- `lang` is the report's language: `zh-TW` when most of the comments you read are Traditional Chinese, otherwise `en`.
## The report
From the repo root, pass the findings file by its path:
```sh
node "<this skill's folder>/scripts/build-report.mjs" "<findings file>"
```
It needs Node 20 or later. The report shows code the script reads from the files itself. When the script reports problems instead, it writes no report; fix the findings file as each message says and run it again.
Finish by giving the user the report path it prints, and remind them to review the rewrites with `git diff`.From GUOJIE-HONG/skills, MIT.
More from GUOJIE-HONG/skills
- 0.589score
ask-jev
GUOJIE-HONG
Let TypeSafe's Jev settle the Open questions your sources leave, and collect the rest into a questionnaire for the person who decides.
SkillAFreshness: A★ 1Claude Code +9 - 0.589score
design-code-implement
GUOJIE-HONG
Turn a settled requirement into at least three grounded implementation directions, then record the chosen one.
SkillAFreshness: A★ 1Claude Code +9 - 0.589score
open-course
GUOJIE-HONG
Turn a topic into a Course project, a Syllabus plus a Handout and Slides per Lesson, where every factual claim cites a Source read for this Course and a point…
SkillAFreshness: A★ 1Claude Code +9 - 0.589score
show-grill-clearly
GUOJIE-HONG
Turn the unresolved decisions of a grilling session into a temporary browser questionnaire, so questions that are hard to follow in chat become clear, while…
SkillAFreshness: A★ 1Claude Code +9 - 0.589score
dont-know-how
GUOJIE-HONG
Find direction for a task you don't know how to start.
SkillAFreshness: A★ 1Claude Code +9 - 0.587score
torture-gently
GUOJIE-HONG
Grill the user about a plan, decision, or idea along grounded, decision-relevant branches.
SkillAFreshness: A★ 1Claude Code +9 - 0.586score
refactor
GUOJIE-HONG
Restructure code under a plan, test-first when behavior changes, against a green baseline when it must stay the same.
SkillAFreshness: A★ 1Claude Code +9 - 0.584score
sib
GUOJIE-HONG
Say it back — restate the user's goal and the problem they are solving in your own words, then wait for them to confirm before any work.
SkillAFreshness: A★ 1Claude Code +9 - 0.584score
cloud-architect
GUOJIE-HONG
Plan how to deploy a repo, to the cloud or on-premises, securely and with good performance, every recommendation backed by evidence.
SkillAFreshness: A★ 1Claude Code +9 - 0.584score
seen-it-before
GUOJIE-HONG
Name the classic problems a hard problem is a variant of, with their known solutions and where the problem departs from them.
SkillAFreshness: A★ 1Claude Code +9 - 0.584score
to-tasks
GUOJIE-HONG
Split each ticket into implementation tasks that write to at most two projects, nested under the ticket.
SkillAFreshness: A★ 1Claude Code +9 - 0.584score
implement-task
GUOJIE-HONG
Execute to-tasks tickets by dependency order, one fresh sub-agent per ticket in its own worktree, independent tickets in parallel, each ticket merged back and…
SkillAFreshness: A★ 1Claude Code +9
Signals
- Stars
- 1
- Forks
- 1
- Open issues
- 1
- Last commit (this folder)
- Oct 21, 2026 (10 days ago)
- Last release
- Oct 23, 2026 (7 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/GUOJIE-HONG/skills
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from GUOJIE-HONG
- 0.589score
open-course
GUOJIE-HONG
Turn a topic into a Course project, a Syllabus plus a Handout and Slides per Lesson, where every factual claim cites a Source read for this Course and a point…
SkillAFreshness: A★ 1Claude Code +9 - 0.589score
show-grill-clearly
GUOJIE-HONG
Turn the unresolved decisions of a grilling session into a temporary browser questionnaire, so questions that are hard to follow in chat become clear, while…
SkillAFreshness: A★ 1Claude Code +9 - 0.589score
design-code-implement
GUOJIE-HONG
Turn a settled requirement into at least three grounded implementation directions, then record the chosen one.
SkillAFreshness: A★ 1Claude Code +9 - 0.589score
dont-know-how
GUOJIE-HONG
Find direction for a task you don't know how to start.
SkillAFreshness: A★ 1Claude Code +9 - 0.589score
ask-jev
GUOJIE-HONG
Let TypeSafe's Jev settle the Open questions your sources leave, and collect the rest into a questionnaire for the person who decides.
SkillAFreshness: A★ 1Claude Code +9 - 0.587score
torture-gently
GUOJIE-HONG
Grill the user about a plan, decision, or idea along grounded, decision-relevant branches.
SkillAFreshness: A★ 1Claude Code +9
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 clean-comments 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.
Cite this page
clean-comments skill. Agent Mart. https://agentmart.dev/skills/guojie-hong-clean-comments (last updated 2026-10-24)