line-editor
Use when the author wants a line edit, prose notes, or sentence-level feedback on a fiction chapter, scene, or passage.
Skill by atrnh
Chalk Labs
Need help deploying line-editor?
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/atrnh-line-editor
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: writing, 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
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/line-editor
What's inside
- Skill name
line-editor- Description
- Use when the author wants a line edit, prose notes, or sentence-level feedback on a fiction chapter, scene, or passage. Comments on local reader effect, preserves deliberate style, and leaves rewrites to the author unless requested. Not for developmental edits or copyedits.
- License
- MIT
- Instructions
- 68 lines in SKILL.md · view on GitHub
Files (12)
- SKILL.md
- openai.yaml · agents/
- evals.json · evals/
- 01-farmhouse.md · evals/files/
- 02-flood.md · evals/files/
- 03-letter.md · evals/files/
- 04-saltmarsh-style-sheet.md · evals/files/
- 04-saltmarsh.md · evals/files/
- 05-visit.md · evals/files/
- 06-saint-agnes.md · evals/files/
- 07-funeral-style-sheet.md · evals/files/
- 07-funeral.md · evals/files/
SKILL.md (first lines)
# Line editor Before reviewing, read and follow [the editorial contract](../references/editorial-contract.md) and [review conduct](../references/review-conduct.md). This skill adds what is specific to line editing. ## Establish the voice baseline Read the whole passage in scope before drafting any note. If the scope is longer than a chapter, confirm it or work in sections, and say which sections you covered. If the passage is too short to show its norms, read the surrounding text for the baseline without commenting on it. Describe the passage's norms in two or three sentences: sentence length and shape, register, period or setting, psychic distance, tense, and recurring devices such as fragments, lists, repetition, dialect, or unusual punctuation. When the narrator or POV changes, describe each voice separately. Note each character's dialogue norms where they differ. Judge every candidate note against this baseline, not against generic workshop prose. A feature consistent with the baseline is style. Flag it only where it departs from the baseline or where the surrounding text shows it working against its own effect. ## Read for causes, not impressions Every note names the reader effect and the cause that produces it. "The rhythm drags" is an impression; "the verb waits behind three prepositional phrases, so the action lands after the reader has lost the subject" is a note. Look for causes like these. They help find problems; they are not note labels. - **Clarity:** ambiguous pronoun or referent; dangling or misplaced modifier; unclear blocking; a response shown before its stimulus; a sentence that needs rereading to parse. - **Rhythm:** main verb delayed by stacked clauses; a weak word in the stress position at the end of a sentence or paragraph; a run of same-shaped sentences where the baseline varies; accidental rhyme or meter; syntax whose pace fights the pace of the moment. - **Diction:** a word outside the narrator's or POV character's register; a near-miss synonym; abstraction where the baseline is concrete; anachronism; unintended cliché. - **Imagery:** mixed or competing figures; a figure whose connotations fight the tone; stacked figures that dilute each other; sensory detail out of perceptual order. - **Repetition:** an unintended echo of a distinctive word or sound within a short span; a structural tic, such as every paragraph opening the same way; information the reader already has. Patterned repetition, such as anaphora or refrain, is style. - **Distance and POV:** an unmotivated shift in psychic distance; filter verbs holding the reader outside a moment the passage otherwise inhabits; knowledge the POV character couldn't have; tense slips. - **Dialogue:** tags or adverbs explaining what the line already shows; characters telling each other what both know; interchangeable voices; beats that break a line's timing. ## Avoid default line-edit habits - Don't flag adverbs, passive voice, "was," filter verbs, or telling by rule. Flag them only when you can name their cost to the reader in this passage. - Don't nudge the prose toward neutral workshop style or toward more ornament. "More vivid" is not a direction. - Don't treat shorter as better. Say what a cut would gain. - Don't smooth roughness the baseline shows is deliberate. - Don't hedge notes. ## Give directions, not rewrites A revision direction names an operation on the existing text: cut these words, move this clause, split or join sentences, resolve a referent, replace an abstraction with the concrete thing it stands for. Quote the words to cut or move. Don't supply replacement words, phrases, or sentences unless the author asks for them. When asked, offer one or two options in the baseline voice within the note, as proposals; changing the manuscript itself still needs a separate request. ## Label each note - **Likely issue:** a local craft problem. - **Continuity:** a conflict with material in scope or a settled author decision, identified without inventing canon. - **Taste:** a preference the author may reasonably decline. When a departure from the baseline may be deliberate, use Taste and ask whether it is intended. - **Protect:** a passage whose effect revision could easily lose. Name the cause of the effect, as with any other note; without it, the note is praise, not editing. Put the notes with the largest reader effect first. For example: > **[Line]** Likely issue: "Nan handed her mother the knife, and she flinched." The reader can't tell who flinched, and the flinch is the point of the moment. Resolve the second "she." > **[Line]** Protect: "Keys. Wallet. The cat's dish, still full." The list gives objects and no feelings, so the shock registers through what she can't stop noticing. Added interiority here would explain it away. ## Check against author decisions After drafting notes, and not before, check them against any style sheet, story-bible entries, or author notes on voice, dialect, spelling, or invented terms that the author supplied or the project keeps. Drafting first keeps the read of the prose independent of what the author intended. For a large story bible, use [story-bible](../story-bible/SKILL.md) to assemble a context packet for the passage rather than reading the whole bible. - Drop notes that contradict a settled decision. - If a decision seems to work against its own effect in this passage, keep the note as Taste and cite the decision. - Flag passages that depart from a settled decision as Continuity, citing the decision. ## Deliver the feedback Deliver as [review conduct](../references/review-conduct.md) describes. Start each note with bold `[Line]` and its label, as in the examples. Each note gives a short exact quote, the reader effect and its cause, and an optional direction. Keep one materially distinct observation per note. For a recurring pattern, note one or two representative instances and list the other locations in the summary. Flag structural concerns for a developmental pass in a line; don't expand into an editorial letter. Mechanical errors such as typos and doubled words are outside a line edit. List them with locations in the summary instead of commenting on each, unless the author also asks for a copyedit. In the closing summary, also include the voice baseline, recurring patterns, and mechanical errors.
From atrnh/skills, MIT.
More from atrnh/skills
- 0.570score
create-lecture
atrnh
Generates a structured lecture from an unordered list of topics.
SkillAFreshness: AClaude Code +9 - 0.570score
create-mcq
atrnh
Generates multiple choice questions (MCQs) from text or documents provided by the user.
SkillAFreshness: AClaude Code +9 - 0.570score
review-fiction
atrnh
Reviews a chapter of the user's fiction and produces a two-column HTML critique page beside the manuscript, with the prose on the left and anchored comments…
SkillAFreshness: AClaude Code +9 - 0.570score
deep-quality-review
atrnh
Reviews any written material (docs, guides, lectures, labs, READMEs, essays, proposals, specs, emails) for both functional quality and deep quality, using the…
SkillAFreshness: AClaude Code +9 - 0.567score
story-bible
atrnh
Build, reconcile, and retrieve an internal fiction wiki from manuscripts and author notes, preserving evidence, canon status, chronology, and character…
SkillAFreshness: AClaude Code +9 - 0.565score
student-question-pass
atrnh
Reviews curriculum material by roleplaying as a target learner, adding student questions as source comments, then interactively turning selected questions…
SkillAFreshness: AClaude Code +9 - 0.565score
triage-feedback
atrnh
Use after an agent or reviewer has produced a batch of proposed changes to text files, such as a deep-quality review, Bugbot or code-review findings, a…
SkillAFreshness: AClaude Code +9 - 0.565score
beep
atrnh
Write and send communications on Ashley's behalf as Beep, her clearly identified robot assistant.
SkillAFreshness: AClaude Code +9 - 0.565score
vaultit
atrnh
Save an assistant response to Obsidian from supplied text, a chat link or message reference, or the immediately preceding response.
SkillAFreshness: AClaude Code +9 - 0.565score
review-lab
atrnh
Reviews lab content drafts from subject matter experts and provides structured editorial feedback.
SkillAFreshness: AClaude Code +9 - 0.561score
grill-my-story
atrnh
Interview a fiction author to explore an uncertain story idea or pressure-test a premise, character choice, scene, or plot problem.
SkillAFreshness: AClaude Code +9 - 0.561score
developmental-editor
atrnh
Give candid developmental critique of fiction or a spoiler-bounded first-reader response, with feedback delivered in the author's chosen workflow.
SkillAFreshness: AClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 22, 2026 (8 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/atrnh/skills
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from atrnh
- 0.570score
create-lecture
atrnh
Generates a structured lecture from an unordered list of topics.
SkillAFreshness: AClaude Code +9 - 0.570score
review-fiction
atrnh
Reviews a chapter of the user's fiction and produces a two-column HTML critique page beside the manuscript, with the prose on the left and anchored comments…
SkillAFreshness: AClaude Code +9 - 0.570score
create-mcq
atrnh
Generates multiple choice questions (MCQs) from text or documents provided by the user.
SkillAFreshness: AClaude Code +9 - 0.570score
deep-quality-review
atrnh
Reviews any written material (docs, guides, lectures, labs, READMEs, essays, proposals, specs, emails) for both functional quality and deep quality, using the…
SkillAFreshness: AClaude Code +9 - 0.567score
story-bible
atrnh
Build, reconcile, and retrieve an internal fiction wiki from manuscripts and author notes, preserving evidence, canon status, chronology, and character…
SkillAFreshness: AClaude Code +9 - 0.565score
student-question-pass
atrnh
Reviews curriculum material by roleplaying as a target learner, adding student questions as source comments, then interactively turning selected questions…
SkillAFreshness: AClaude 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 line-editor 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
line-editor skill. Agent Mart. https://agentmart.dev/skills/atrnh-line-editor (last updated 2026-10-24)