Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

Back to skills

Scope

ASecurity

Establish or confirm a piece of writing's audience, tone, and purpose before reviewing it. Phase 1 of the writing pipeline — run standalone or via writing:full.

8 stars
0 votes
0 copies
0 views
Added 9/20/2026
ai-agentsgogit

Security Analysis

A100/100

Scanned 9/20/2026

Install to Claude Code

$npx -y skills add tstapler/dotfiles --skill scope --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Scope?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Scope
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/tstapler-scope/badge)](https://www.skillsdirectory.com/skills/tstapler-scope)

More formats (shields.io, HTML) on the badges page.

Download Zip
Files
SKILL.md
---
description: Establish or confirm a piece of writing's audience, tone, and purpose before reviewing it. Phase 1 of the writing pipeline — run standalone or via writing:full.
---

# writing:scope

Every other check in this pipeline (`writing:tone`, `design-doc-review:readability`, `technical-writing-coach`, `writing-humanize`) needs to know who the text is for and what register it should hit. Getting that wrong produces confident, precise feedback aimed at the wrong target — a PR description reviewed as if it were a design doc, or a blog post held to a design doc's decision-focus standard. This phase exists so every later check shares one answer instead of each guessing its own.

**Target**: {{args}} — a file path, or text already in context.

## Step 1 — Detect what you can before asking

Read the target and check for signals that already answer some of this:

- **Design doc / RFC**: a Status/Tracking/Decision-requested header, or it lives under a `design-docs`-style repo with its own house-style `CLAUDE.md` (check for one). If so, audience and tone are largely fixed by that repo's convention — don't re-litigate them, just confirm.
- **PR description / commit message**: git context (is this text going into `gh pr create --body` or a commit?) — audience is reviewers of this specific change, tone is terse and evidence-first per this repo's own CLAUDE.md if one exists.
- **Blog post / external-facing**: no internal jargon assumed, audience is whoever the platform's readership is.
- **Personal note / journal**: audience is future-you; tone should stay in the author's own voice — this is the one case where `writing-humanize`'s AI-tell removal matters most, since a note in someone else's cadence is actively worse than one in the author's.

If the type is unambiguous from context, state your read and move to Step 3 without asking. Only ask when it's genuinely unclear or when a decision only the author can make is in play (see CLAUDE.md's escalation rule).

## Step 2 — Ask what you can't detect

When ambiguous, ask in one pass (not one question at a time):

```
questions:
  - question: "Who's the primary audience for this?"
    header: "Audience"
    options:
      - label: "Expert peers in this domain"
        description: "Can assume shared jargon and context"
      - label: "Cross-functional / mixed technical background"
        description: "Define terms on first use, less assumed context"
      - label: "External / general readership"
        description: "No internal jargon, no assumed org context"
  - question: "What register should this land in?"
    header: "Tone"
    options:
      - label: "Formal / decision-document"
        description: "Structured, evidence-cited, matches this repo's house style if one exists"
      - label: "Direct / conversational"
        description: "Plain sentences, contractions fine, minimal hedging"
      - label: "Match my own voice"
        description: "Preserve the author's existing cadence — flag AI-tells (writing-humanize) more aggressively"
```

Never infer "formal" as the default just because the source reads dense — that density might itself be the problem this whole workflow was invoked to fix. Ask rather than assume when the current draft's register and the target register might be the same mistake.

## Step 3 — Record the scope

Hold the answer as: `{audience, tone, doc_type, purpose}`. Don't write a sidecar file for a one-off review — that's persistence for something used once in this session. Only write `<target>.writing-scope.md` if the user says this doc will go through multiple review rounds across sessions and wants the scope to persist; otherwise pass it directly into the next phase's agent prompts.

State the resolved scope back to the user in one line before proceeding: `Scope: <doc_type>, audience=<audience>, tone=<tone>.`

## Handoff

`writing:full` calls this phase first and threads its output into every check agent's prompt in `writing:tone` and (if the target isn't already a design doc using `design-doc-review:review`'s own checks) `technical-writing-coach`.

Attribution

tstaplertstapler
View sourceMore from tstapler →
SSkills DirectorySkills Directory

Your tool, in front of Claude Code builders.

3 founder slots · $299/mo · GSC-verified traffic · sponsors can never buy grades.

See placements

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Your tool, in front of Claude Code builders.

3 founder slots · $299/mo · GSC-verified traffic · sponsors can never buy grades.

See placements

Related Skills

Caveman

Ultra-compressed communication mode. Cuts token usage ~75% by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, wenyan-ultra. Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens", "be brief", or invokes /caveman. Also auto-triggers when token efficiency is requested.

1023331 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

686011 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3331 votes

catchup

Recovers prior coding-agent session context by running `catchup <agent> --since-compact`, which extracts a clean summary of a previous Codex, Claude Code, Antigravity, OpenCode, or Pi Agent session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", or asks to recover/summarize a previous session before continuing. Do NOT use for the current conversation, git history, or any non-agent log.

611 votes

math-skill

A comprehensive mathematical reasoning skill for AI assistants — handles arithmetic to research-level problems with rigorous step-by-step reasoning, systematic verification, and transparent uncertainty handling

381 votes
View all in ai-agents →