Marketplace for trusted AI agent skills

Skill Detail

Run Instagram caption and carousel drafting with Instagram Skills

Transform campaign notes into Instagram captions and slide-by-slide carousel drafts, then review factual claims, voice, media order, and an approval card.

Content Writing & SEOMulti-Framework
Content Writing & SEO Multi-Framework Security Reviewed
⭐ 455 GitHub stars
COPY SKILL INSTRUCTIONS (OPTIONAL)
npx skills add agentskillexchange/skills --skill plan-and-draft-instagram-captions-and-carousels-with-instagram-skills Copy
Uses the third-party skills CLI, not an ASE-owned installer. Check your agent’s compatibility. This copies instructions; complete the upstream tool setup below separately.
View source Documentation
At a glance
Tools required
SKILL.md-compatible agent (Claude Code, Codex, or explicit Markdown context loading); Git for clone setup; Python and repository dependencies only for optional integrations; optional Publora credentials and supplied media for publishing
Install & setup
Claude Code: run /plugin marketplace add sergebulaev/instagram-skills, then /plugin install instagram-skills@instagram-skills. For another SKILL.md-compatible runtime, clone https://github.com/sergebulaev/instagram-skills.git and explicitly load SKILL.md plus the relevant skills/*/SKILL.md. Draft-only use needs no API keys. For optional Python integrations, create an isolated virtual environment in the cloned repository and run python -m pip install -r requirements.txt. Configure optional service credentials through the documented .env.example; publishing requires a connected Instagram Business or Creator account.
Author
Creative Content Crafts (sergebulaev)
Publisher
Community
Last updated
Oct 16, 2026
Quick brief

Run an Instagram content drafting workflow with the Instagram Skills repository. Give the agent a campaign brief, audience, goal, source facts, voice examples, and any supplied media. Load the root SKILL.md router, then the focused ig-caption-writer or ig-carousel-planner skill. For a carousel, choose an appropriate formula, set a justified slide count, write the opening hook, map the value slides, write the payoff slide, and draft the supporting caption.

How it works

What this skill actually does

Invoke this workflow when you want an agent to transform actual campaign notes into reviewable Instagram copy before posting. For example: turn supplied launch notes into a six-slide carousel with slide text, a caption, and one call to action. Review each factual claim against the inputs, apply the supplied voice profile, and return an approval card with the full copy, hashtag set, slide order, primary goal, and missing media. Upstream explicitly forbids inventing numbers, dates, client names, and results.

The scope boundary is one caption or carousel preparation task. Outputs are copy, an ordered slide outline, and a media checklist. Draft-only operation requires no API credentials and ends with material for manual posting. Related LinkedIn skills cover another platform; read-only social research and download tools do not replace Instagram slide architecture and campaign-to-copy preparation. Upstream marketing and ranking heuristics are guidance, not independently verified platform guarantees.

Install in Claude Code with /plugin marketplace add sergebulaev/instagram-skills and /plugin install instagram-skills@instagram-skills. Alternatively, clone the exact repository and explicitly load its root and focused skill files in a compatible agent. The README documents Claude Code, Codex, OpenClaw, Hermes, and manual Markdown invocation. This supports Multi-Framework through actual skill invocation paths.

Optional publishing is a separate action requiring explicit approval of copy, media, account, and timing. Upstream documents Publora configuration, Python dependencies, a connected Business or Creator account, and a draft/upload/schedule sequence. The carousel skill documents 2-10 ordered images for that route. This approved workflow requires no publishing integration. Approval cards are instruction-level conventions, not verified runtime consent enforcement. Repository metadata and skill instructions were inspected; no installation, media upload, paid integration, or publication was performed. Metadata verified does not mean security reviewed.