article-outline
Outline before draft — the text's purpose, its Diátaxis type, audience, style forks and sections, agreed before a word of body text. Use when a text (article, doc page, guide, report) is being planned, or when a draft is about to start without an agreed type and reader.
- License 许可
- MIT
- 所属仓库
- muratovv/ai-hats
- 最近更新
- 2026-11-05
- Stars
- ★ 3
- Forks
- 1
01 前言 Front Matter
- name
- article-outline
- license
- MIT
- description
- Outline before draft — the text's purpose, its Diátaxis type, audience, style forks and sections, agreed before a word of body text. Use when a text (article, doc page, guide, report) is being planned, or when a draft is about to start without an agreed type and reader.
03 SKILL.md 正文
Article Outline
Settle what the text is, whom it serves and how it will be styled before drafting. The outline is the text's plan; the draft follows it.
When to Use
- A text is being planned, or a draft is about to start and nobody has said
what type it is and who reads it. A draft that already exists is
prose-review's job. - With a task card, the outline is its own
## Outlinesection in the plan, written and agreed before the draft starts. It ADDS that section — it never fills one that already has an owner (Requirements, Steps): an outline is the text's own shape, not the task's requirements or its work breakdown. Without a card it goes to chat and is agreed there before the draft.
Procedure
-
Purpose. One sentence: who reads this and what they leave with. It becomes the text's lead.
-
Type — one Diátaxis quadrant, held for the whole text:
Quadrant The reader wants to Shape Tutorial learn by doing ordered steps, one working result, no options How-to solve a task they have goal → steps → verification; assumes competence Reference look something up one structure per item, complete, no narrative Explanation understand why argument: claim → reasons → consequences A text serving two quadrants is two texts, or one with a named split.
-
Audience. One line: what they already know, what they don't, what they came to do. It decides which terms get defined and which steps get skipped.
-
Style forks — each settled as a line in the plan, before drafting: reference format (inline / numbered / footnote); voice (first person / impersonal, imperative / descriptive); code blocks (the reader runs / the agent ran / illustration only); links (first mention / every mention); terms source (a glossary, or "this text defines its own").
-
Sections. One heading per reader gain, one line each: what the reader gets here. No gain, no section. In a tutorial the sections are the steps; in a reference they are the items.
-
Agree the outline with the author. Body text starts after that.
Completion
- Purpose sentence, one quadrant, audience line, style forks and sections
with gains are written into the plan's
## Outline(or chat) and agreed. - Validation — RED: asked for "an article about worktrees", the agent starts drafting a tutorial-explanation hybrid for an unnamed reader. GREEN: it returns the outline first and asks for one decision — the quadrant.
Anti-Patterns
- Drafting first and outlining backwards.
- "Overview" as a section's gain.
- Two quadrants in one text without a named split.
- Style forks settled silently — each one costs a review round later.
04 同仓库更多 Skills
tool-evaluation-protocol
Time-bounded protocol for evaluating new tools, libraries, or methodologies before adoption. Use when evaluating a candidate dependency for the project, researching a methodology (e.g. ML/LLM frameworks, testing approaches), or following up on a hype-driven recommendation.
systematic-debugging
4-phase bug-fix protocol (evidence, pattern, hypothesis, verify). Use when investigating any error, failure, or unexpected behavior, handling bug reports from users or monitoring, or diagnosing test failures with non-obvious causes.
skill-template
Canonical template and validation guide for ai-hats skills. Use when creating a new skill, reviewing an existing one for structural compliance, or deciding which pattern fits a new behavior.
skill-optimization
Audit and refactor library components to eliminate redundancy and staleness. Use during periodic library maintenance, after adding multiple new skills or rules, when a role's token budget is too high, or after retrospectives identify component issues.
silent-fallback-gate
PreToolUse hook — refuses an Edit or Write that adds an inert broad except. Use when a Python edit was blocked for a silent fallback.
retro-to-framework
Convert an observation about agent behavior — from a retrospective or from a single incident — into a framework-level change (hook, rule, skill, skill update). Use after a retrospective identifies problems that are not project-specific, after an incident where an agent misapplied a component it had read, when CLAUDE.md band-aids accumulate (more than 3 per-project rules that could be generic), or when the same problem recurs across multiple projects.