prompt-writer SKILL.md skillF

prompt-writer from anilcancakir/claude-code is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add anilcancakir/claude-code/prompt-writer@git:20261011.47f52bb

Pins git:20261011.47f52bb in mdr.lock. Pin the fingerprint instead.

Audit grade
F, 16 of 17 checks passed
Kind
SKILL.md skill
Repository
anilcancakir/claude-code
Path
plugins/ac/skills/prompt-writer/SKILL.md
Versions recorded
9
Latest version
git:20261011.47f52bb, committed Oct 11, 2026
First version recorded
Jul 31, 2026
Size
25,804 bytes
Repository stars
3
License
MIT

Write or audit a prompt another Claude will run: system prompts, subagent briefs, skill, command and agent bodies, CLAUDE.md and rules.

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# Prompt Writer

You are about to write or edit a prompt another Claude will execute. This skill is the playbook: rules, architecture, snippets, and worked examples for producing a high-signal prompt on the first try. Primary target is Claude Opus 5.5 (`model: opus` on Claude Code 2.1.280), whose documented baseline is Opus 5; the same shapes work on Sonnet 5.5 (`model: sonnet` on 2.1.284) at lower cost and on Fable above it.

Skim this body, jump to the reference that matches the task, fill in the template, validate against the checklist. The body carries the workflow; the references in `${CLAUDE_SKILL_DIR}/references/` carry the depth.

## Ground before you write

Do not author a prompt from your own built-in knowledge alone. Parametric knowledge is frozen at the training cutoff; the world the prompt runs in is not. Two grounding passes come before the first line, every time.
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add anilcancakir/claude-code/prompt-writer@sha256:b176b3f923bf64cb

The line at the top of this page pins the label git:20261011.47f52bb, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_jtehk2qeppqyijgx.svg)](https://markdownregistry.com/a/art_jtehk2qeppqyijgx)

3 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 9 versions of this file, between Jul 31, 2026 and Oct 11, 2026, about one every 9 days.

versioncommittedcommitsizeaudit
git:20261011.47f52bb latest2026-10-11 47f52bb 25,804 BF view · diff
git:20260928.7f5fecd2026-09-28 7f5fecd 25,192 BF view · diff
git:20260923.539bf3d2026-09-23 539bf3d 23,861 BF view · diff
git:20260923.e35d9172026-09-23 e35d917 23,778 BF view · diff
git:20260904.925ce452026-09-04 925ce45 23,030 BF view · diff
git:20260903.55c81132026-09-03 55c8113 23,264 BF view · diff
git:20260903.ada2f102026-09-03 ada2f10 23,685 BF view · diff
git:20260902.b711a792026-09-02 b711a79 23,049 BF view · diff
git:20260731.a3791442026-07-31 a379144 23,350 BF view

Audit of the latest version

F  16 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • FAIL, critical: No instruction hidden inside an HTML comment (matched: <!-- SYSTEM PROMPT -->)
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (25804 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Starting a new one? Use a template for SKILL.md on mdpantry, which passes every one of these checks and says what to change in it.

Source

GitHub

anilcancakir/claude-code · 3 stars · license MIT · pushed 2026-10-19 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_jtehk2qeppqyijgx
GET https://markdownregistry.com/api/v1/resolve?ref=anilcancakir/claude-code/prompt-writer
GET https://markdownregistry.com/api/v1/blob/b176b3f923bf64cbf8fcf8e2d13e046b693b2e1e1959b53c10e1ed0c5f82030d

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

More from anilcancakir/claude-code

agent-creator SKILL.md skill
anilcancakir/claude-code · plugins/ac/skills/agent-creator/SKILL.md · Create or audit a Claude Code subagent: tools, model and effort, permissions, memory, isolation, and the system-prompt…
git:20261016.83d4487 · audit A · 3 stars
auto SKILL.md skill
anilcancakir/claude-code · plugins/ac/skills/auto/SKILL.md · Carry a request through plan, execute and a final verification on a branch, without supervision, or take over the work…
git:20261017.219ebfa · audit A · 3 stars
claude-md-rules-creator SKILL.md skill
anilcancakir/claude-code · plugins/ac/skills/claude-md-rules-creator/SKILL.md · Write or audit CLAUDE.md, CLAUDE.local.md and .claude/rules files: scope, paths, imports, size, and why a rule is…
git:20261016.83d4487 · audit F · 3 stars
command-creator SKILL.md skill
anilcancakir/claude-code · plugins/ac/skills/command-creator/SKILL.md · Create or audit a Claude Code slash command: arguments, shell-injected context, phases, and approval gates.
git:20261016.b354c89 · audit A · 3 stars
coordinate SKILL.md skill
anilcancakir/claude-code · plugins/ac/skills/coordinate/SKILL.md · Turn a large request into a reviewed roadmap of bundles and drive them through parallel ac:lane worktrees.
git:20261019.6bdb254 · audit A · 3 stars
execute SKILL.md skill
anilcancakir/claude-code · plugins/ac/skills/execute/SKILL.md · Run an approved ac plan to completion, wave by wave, with cost-tiered workers, per-wave verification and a final review.
git:20261019.82eee14 · audit B · 3 stars
git-master SKILL.md skill
anilcancakir/claude-code · plugins/ac/skills/git-master/SKILL.md · Make atomic commits in the repo's own style, rebase or squash, and trace history with blame, bisect and pickaxe.
git:20261016.b354c89 · audit A · 3 stars
plan SKILL.md skill
anilcancakir/claude-code · plugins/ac/skills/plan/SKILL.md · Plan a change interactively: research in parallel, ask the decisions only you can make, then write a tiered plan…
git:20261019.6bdb254 · audit B · 3 stars
qa SKILL.md skill
anilcancakir/claude-code · plugins/ac/skills/qa/SKILL.md · Test a session's or a plan's work the way a QA engineer would: derive cases from the requirement and the diff, run them…
git:20261017.48efb43 · audit A · 3 stars
skill-creator SKILL.md skill
anilcancakir/claude-code · plugins/ac/skills/skill-creator/SKILL.md · Create or audit a Claude Code skill: frontmatter, layout, scope, invocation, and bundled references or scripts.
git:20261016.b354c89 · audit A · 3 stars

Every file in anilcancakir/claude-code

Browse by kind, by grade F, or by owner.