agentleFS
MaGe1993MaGe1993/Claude-Code-Base-Harness/.claude/skills/proofread/SKILL.md

proofread

Skill1 starsChanged 5 months agoMITView on GitHub

Verify a finished deliverable before it ships — numbers consistent, names and proper nouns right, facts accurate, internally consistent, logically sound. Read-only; returns severity findings.

26 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use humanize, deploy and 2 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • eherrerosjSkill

    humanize

    Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive 'Signs of AI writing' guide. Detects and fixes patterns including: inflated symbolism, promotional language, superficial -ing analyses, vague attributions, em dash overuse, rule of three, AI vocabulary words, negative parallelisms, and excessive conjunctive phrases. Triggers on: 'humanize', 'make this sound human', 'remove AI writing', 'de-AI', 'sound more natural', 'less robotic', 'too AI-sounding', or when reviewing text that reads like AI output.

    Also in 3 projects that use this

  • VercelSkill

    deploy

    Deploy applications to Vercel. Use when the user says "deploy", "deploy to Vercel", "push to production", "deploy my app", or "go live".

    Also in 3 projects that use this

  • Microsoft AzureSkill

    verify

    End-to-end app testing pipeline for browser-based web apps on cloud browsers via Azure Playwright Workspaces: verify the live app, author tests for working flows, heal failing tests, then re-run with video and produce one consolidated report. Three phases (app verification, test authoring, test healing) plus a scoped final sanity run; each phase emits a sub-report. Use when asked to "verify my app", "make sure my app works and has good test coverage", "do a full QA pass", "add tests where I'm missing coverage", "fix failing tests and add new ones", "smoke test before deploy", "QA my deployment", or any combined verify-and-test ask. Requires Azure CLI (`az login`) and a Playwright Workspace resource ID — bootstraps configs on first invocation. Browser-based web apps with Playwright only — not for unit/API tests, native mobile/desktop, load/perf, or full accessibility audits.

    Also in 3 projects that use this

  • SentrySkill

    review

    Fixture review skill.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 20 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read-only. Produce findings.

    line 27
  2. Distinctive:

    Verify against the source when one is available (the data, inputs, or brief the deliverable is based on).

    line 29
  3. Distinctive:

    When no source is available: check internal consistency and plausibility, and flag every unverifiable figure or claim explicitly rather …

    line 31
  4. Distinctive:

    Never soften a finding. A transposed figure or a wrong client name is a critical defect, not a nitpick.

    line 33

“Common” is measured: kept by at least 6 in 10 of the 20 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Proofread

1§ Purpose

The process schema for proofing a finished deliverable — a report, proposal, slide deck, memo, or any content destined for a client or external recipient. Verify that the content is correct and ready: figures, names, facts, consistency, and logic. Read-only: produce findings, never modify.

This checks the result, not the process or the code. Technical quality of source code or harness artifacts, and security, are out of scope — this skill is content correctness only.

It is loaded by whoever needs it — typically the ag-cqo agent as a final deliverable gate, or the main session for an inline proof.

2§ Use when

  • A deliverable is finished and about to be sent to a client or published.
  • User asks to "proofread", "fact-check", "check the numbers", "final review before sending", or "is this ready to go out?".
  • Any finished document needs a correctness and consistency pass.

3§ Hard constraints

  • Read-only. Produce findings; never edit the deliverable.
  • Cite an exact location for every finding: page, section, heading, table, or figure.
  • Verify against the source when one is available (the data, inputs, or brief the deliverable is based on). Read it before judging a figure or claim correct.
  • When no source is available: check internal consistency and plausibility, and flag every unverifiable figure or claim explicitly rather than passing it silently.
  • Never soften a finding. A transposed figure or a wrong client name is a critical defect, not a nitpick.
  • Distinguish confirmed errors from suspected ones.

4§ Dimensions

Apply all that the deliverable contains.

  • P1 Numbers — every figure correct and internally consistent: totals add up, subtotals match the total, the same quantity is stated identically everywhere it appears, percentages sum as expected, units and currencies correct, dates and amounts not transposed, a number in the text matches the number in the table or chart it refers to, cross-references ("see Table 3") point to the right object.
  • P2 Names and proper nouns — person, company, product, and place names spelled and capitalized correctly and consistently throughout; the client's name exactly right everywhere; trademarks per assets/proper-nouns.md; titles and roles correct.
  • P3 Facts — every checkable claim matches its source; quoted figures match the underlying data; citations and references resolve; no claim contradicts the provided inputs.
  • P4 Internal consistency — no statement contradicts another; terminology used consistently; the executive summary matches the body; numbers in text match numbers in tables and charts; recommendations match the findings they rest on.
  • P5 Logic — conclusions follow from the stated evidence; no non-sequiturs; assumptions are stated; no gap where a claim depends on something not shown; recommendations are consistent with the analysis.
  • P6 Completeness and language — no leftover placeholders, TODOs, tracked changes, or draft markers; every promised section present; spelling and grammar correct; tone and register fit the recipient.

5§ Severity levels

  • critical — would mislead or embarrass the recipient: wrong figures, wrong client or person name, a factual error, a self-contradiction, a logical break. Must be fixed before sending.
  • major — a real error that should be fixed before sending but is unlikely to mislead on its own.
  • minor — polish: wording, minor inconsistency, formatting.
  • info — observation; no action required.

6§ Workflow

  1. Identify the deliverable and, if available, its source (data, brief, prior version). Read both.
  2. Apply P1–P6. For every figure and factual claim, trace it to the source or mark it unverifiable.
  3. Record each finding: dimension, severity, exact location, what is wrong, the corrected value or the fix, and whether it is confirmed or suspected.
  4. Group by severity: critical → major → minor → info.
  5. Deliver findings. State what was checked, what source was used, and what could not be verified.

7§ Output contract

Must include:

  • Scope statement: the deliverable checked, the source used (or "no source available"), dimensions applied
  • Findings table: dimension | severity | location | issue | correction | confirmed/suspected
  • Summary: count per severity; the unverifiable items listed explicitly

Must not include:

  • Edits to the deliverable
  • Findings without an exact location
  • Silent passing of an unverifiable figure or claim

This skill produces findings only. The ship/no-ship stage-gate decision and its format are the calling agent's reporting standard (see ag-cqo), not part of this skill.

8§ Associated documents

  • [assets/proper-nouns.md] — trademark and product-name capitalization reference for P2.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.