Skip to content
Development
Skill

/create

Interactive outline co-creation for developer advocacy content — collects format, audience profile, story arc (Problem→Journey→Insight→Action), and voice/tone; detects out-of-scope requests (FAQs, comparison tables); surfaces conflicts between user brief and audience needs.

From plugin
ai-rig
3148 skills16 agents2 MCP

Claude Code

$ npx -y skills add Borda/AI-Rig --skill create --agent claude-code

SKILL.md

create.SKILL.md
name: create
description: 'Interactive outline co-creation for developer advocacy content — collects format, audience profile, story arc (Problem→Journey→Insight→Action), and voice/tone; detects out-of-scope requests (FAQs, comparison tables); surfaces conflicts between user brief and audience needs. Writes approved outline to .plans/content/<slug>-outline.md for foundry:creator to execute. Use when starting a blog post, Marp slide deck, social thread, talk abstract, lightning talk, animated SVG, or short video (HyperFrames HTML composition; render stays user-run). SKIP: idea generation before an outline exists (use foundry:brainstorm — brainstorm explores directions into a tree; create co-creates the outline, foundry:creator writes the finished artifact).'
argument-hint: '[topic]'
disable-model-invocation: true
allowed-tools: Read, Write, Bash, TaskCreate, TaskUpdate, TaskList, AskUserQuestion, Agent, Skill
effort: medium

<objective>

Story arc four-beat: Problem → Journey → Insight → Action.

NOT for: implementation, code gen, README writing (use `foundry:doc-scribe`), structured ref docs (FAQs, comparison tables — use `foundry:doc-scribe`), media processing of existing files (download, transcode, trim, transcribe — user runs `ffmpeg` / Whisper / `yt-dlp`).

</objective>

<inputs>

  • **$ARGUMENTS**: optional — topic or goal, any form; one sentence enough. Format hints accepted ("a blog post about…", "talk abstract for…").

</inputs>

<workflow>

**Task hygiene** — task tools may be deferred; load before first use: `ToolSearch(query="select:TaskList,TaskCreate,TaskUpdate,TaskGet", max_results=4)`. Call `TaskList` first and triage each task it returns: `completed` if work clearly done, `deleted` if orphaned, keep `in_progress` only if genuinely continuing. Never spend a turn on bookkeeping alone — every `TaskCreate`/`TaskUpdate` ships in the same response as the next substantive tool call; one exception, `TaskUpdate(completed)` immediately before a long output block (`rules/task-lifecycle.md`).

**Task tracking**: TaskCreate all steps in the same response as the first real tool call — never a response of bookkeeping alone.

Step 1 — Parse topic and out-of-scope detection

  • $ARGUMENTS provided: extract topic; note embedded format hint.
  • No $ARGUMENTS: AskUserQuestion — "What are you trying to write about, and for whom?" (free text). After answer, re-check against out-of-scope conditions: answer describes FAQs, comparison tables, feature matrices, README content, or docstrings: stop. Respond: "This format doesn't fit a narrative arc — use `foundry:doc-scribe` for structured reference content." No further steps.
  • Out-of-scope gate (when $ARGUMENTS provided): brief describes FAQs, comparison tables, feature matrices, or ref docs: stop. Respond: "This format doesn't fit a narrative arc — use `foundry:doc-scribe` for structured reference content." No further steps.
  • Media-processing gate (both inputs — $ARGUMENTS and the opener answer): brief only asks to download, transcode, trim, compress, caption or transcribe existing media, no new narrative artifact: stop. Respond: "This is media processing, not a narrative artifact — run `ffmpeg` / Whisper / `yt-dlp` directly; `/foundry:create` authors new content only." No further steps. Source media for a new artifact (own talk recording, demo clip) stays in scope: record its path in outline `## Constraints` as context.

Step 2 — Format and audience (1 AskUserQuestion call — 2 questions; +1 follow-up call when Q1 = (d))

**Format + audience** (single AskUserQuestion call with 2 questions; AskUserQuestion allows at most 4 options per question, so the four short talk and visual formats share slot (d)):

  • Q1 "What content format?": (a) blog post · (b) conference / meetup talk with Marp slide deck ★ · (c) social thread (X/LinkedIn) · (d) short talk or visual — talk abstract, lightning talk, animated SVG, short video
  • Q2 "Who is the audience?": (a) beginners — new to problem space ★ · (b) intermediate — familiar with basics, seeking depth · (c) expert — know landscape, want novel insight · (d) describe your own profile

Q1 = (d) → one follow-up AskUserQuestion call, 1 question "Which format?": (a) talk abstract (CFP submission) · (b) lightning talk (5–10 min) · (c) animated SVG (looping explainer graphic, embeddable via `<img>`) · (d) short video (≤60 s, HyperFrames HTML composition). No ★ — pure user choice. All seven formats stay reachable.

After answer: restate one sentence covering format + audience ("Got it — a [format] for [audience description].").

Step 3 — Arc construction and conflict check

Propose four-beat arc from topic + audience:

  • **Problem**: concrete opening hook — specific pain or question, not generic
  • **Journey**: 3–5 key points (what tried, what failed, what arc covers)
  • **Insight**: core "aha" framed for stated audience level — name directly
  • **Action**: specific next step for audience
  • **Visual formats** (animated SVG, short video): beats map to timed segments (SVG) or scenes (video); Journey 1–3 points max; Action = closing frame/card

**Out-of-scope re-check (re-apply the Step 1 test to the constructed arc)**: the Step 1 gate tests the brief; this one tests what the arc turned out to be. If the Journey beat resolves mainly to a feature-by-feature comparison, an options matrix, a question-and-answer list, or reference material with no narrative through-line — i.e. the beats only hold together as a table or list — stop here, before writing any outline. Respond: "The arc for this topic resolves to structured reference content rather than a narrative — use `foundry:doc-scribe` instead." Do not write the outline file and do not offer the Step 4 generation gate. A topic that merely *mentions* alternatives still passes: the test is whether the Insight beat names a single transferable idea, or just summarises the comparison.

**Editorial conflict check**: brief implies expert audience but topic introductory,

Read more
@borda
Ships withai-rig

Practical agent workflows for Python, ML, and open-source maintenance. AI-Rig turns recurring work—scoping a change, reproducing a bug, reviewing a pull request, running an experiment, or checking release readiness—into explicit workflows with specialist

3148 skills16 agents2 MCP
Get the whole plugin

Stars on GitHub

Momentum

stars in total
31
Star history of Borda/AI-Rig on GitHub, a chart by star-history.com

Other skills on ai-rig.

debug

debug

Investigation-first debugging — gather evidence, form confirmed root-cause hypothesis, hand off to fix mode with diagnosis file.…

@borda@bordaInstall Skill
feature

feature

TDD-first feature development — crystallise API as a demo test, drive implementation to pass it, run quality stack and a bounded…

@borda@bordaInstall Skill
fix

fix

Reproduce-first bug resolution — capture bug in failing regression test, apply minimal fix, run quality stack and a bounded…

@borda@bordaInstall Skill
plan

plan

Analysis-only planning — classify and scope a task without writing code; outputs a structured plan to .plans/active/. TRIGGER…

@borda@bordaInstall Skill
refactor

refactor

Test-first refactoring — audit coverage, add characterization tests, apply changes with safety net, run quality stack and a…

@borda@bordaInstall Skill
review

review

Multi-agent code review of local Python files, directories, or the current git diff covering architecture, tests, performance,…

@borda@bordaInstall Skill

More skills from other plugins.

brainstorming
Auto-invoked

brainstorming

You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or…

@obra@obraInstall Skill
ask-matt

ask-matt

Ask which skill or flow fits your situation. A router over the skills in this repo.

skills2w309k
@mattpocock@mattpocockInstall Skill
accessibility
Auto-invoked

accessibility

Design, implement, and audit accessible UI to WCAG 2.2 Level AA across Web, iOS, and Android — semantic ARIA roles and labels,…

ecc2w293k
@affaan-m@affaan-mInstall Skill
compiler-commit

compiler-commit

Use when you want to verify compiler changes and commit with the correct convention. Runs tests, lint, and format, then commits…

react5d252k
@facebook@facebookInstall Skill

Questions about create.

Which coding agents can use create?
create installs in Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, OpenCode and Windsurf with the skills CLI, using each agent's flag shown on this page. Flowy's auto-invocation, which fires a skill when your prompt matches, runs in Claude Code only.