← Back to Skills Marketplace
112
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install veezvg-episode-to-instagram
Description
Full pipeline to turn a video podcast episode into Instagram content (carousel posts, quote cards, and Reels). Transcribes the episode, extracts key quotes w...
Usage Guidance
This skill appears to implement the described pipeline, but there are a few red flags to consider before installing or running it:
- Metadata mismatch: The SKILL.md (and scripts) require ffmpeg and an OPENAI_API_KEY (Whisper transcription), and optionally a Replicate key and a logged-in OpenClaw browser Instagram session — but the registry metadata declares none of these. Treat the skill as incomplete until metadata is corrected.
- Credential scope: The transcription step will send audio to api.openai.com (transcribe.sh uses curl with $OPENAI_API_KEY). Only provide an API key you trust to use for this purpose. The browser automation will operate using an Instagram session that must already be logged into the OpenClaw browser profile — that session can post to your account, so only run the scripts with an account you control and after reviewing the automation flow.
- Slack/approval mismatch: SKILL.md says generated slides should be sent via Slack for approval, but the repository contains no Slack integration. Plan how approvals will be handled (manual review, separate Slack skill, or local copy) before allowing automatic posting.
- Installation: npm install may build native modules (canvas) and run prebuild installers; do this in a controlled environment. Playwright is expected from the OpenClaw install or system; if missing, posting scripts will fail.
- Safety/operation: Review scripts that run shell commands (post-to-instagram uses execSync) and confirm you understand what will be copied to /tmp/openclaw/uploads. Consider running the pipeline in a sandboxed environment or on a throwaway Instagram account first.
If you plan to proceed, ask the skill author to: (1) update the registry metadata to declare required env vars/binaries, (2) document how approvals are delivered (Slack or other), and (3) optionally implement safer failure-handling and clearer instructions for installing Playwright and native deps.
Capability Analysis
Type: OpenClaw Skill
Name: veezvg-episode-to-instagram
Version: 1.0.1
The skill bundle provides a legitimate end-to-end pipeline for converting video episodes into Instagram content using ffmpeg, OpenAI's Whisper API, and Playwright-based browser automation. The scripts (transcribe.sh, post-to-instagram.js, etc.) are well-documented, align strictly with the stated purpose in SKILL.md, and include safety features such as mandatory user approval steps and preview screenshots before publishing. No evidence of data exfiltration, malicious persistence, or harmful prompt injection was found.
Capability Tags
Capability Assessment
Purpose & Capability
The SKILL.md and code show a legitimate need for ffmpeg, an OpenAI API key (Whisper transcription), and an OpenClaw/browser session for Instagram posting. However the registry metadata declares no required env vars or binaries — a clear mismatch. The skill also mentions optional Replicate usage and Slack approvals in the workflow, but those integrations are not declared in metadata and are not implemented in the codebase (no Slack client or Replicate calls present).
Instruction Scope
The runtime instructions correctly describe the end-to-end pipeline. The scripts perform file I/O, local ffmpeg/ffprobe operations, call OpenAI's API via curl (transcribe.sh), and drive a logged-in Instagram session through the OpenClaw browser using Playwright/CDP. The agent instructions require sending previews/approval via Slack, but the repo contains no Slack integration — this means the agent or user must supply that channel separately. The instructions do not instruct reading unrelated system secrets, but they do require access to a logged-in browser profile (which gives access to the user's Instagram session).
Install Mechanism
No install spec is provided in the registry (instruction-only), but package.json lists a native dependency (canvas) that requires npm install and native build/prebuild tooling. Installing may trigger native build steps (prebuild-install) and fetches from npm; that's expected but worth noting. Playwright is not bundled — the Playwright helper expects playwright-core to be available from an OpenClaw installation or global node_modules, which may lead to runtime failures if environment not prepared.
Credentials
SKILL.md explicitly requires OPENAI_API_KEY (and optionally REPLICATE_API_TOKEN) and ffmpeg, plus a logged-in OpenClaw browser Instagram session. The registry metadata, however, lists no required env vars or binaries. The discrepancy is worrying because the transcribe script uses the OPENAI_API_KEY environment variable directly; that credential is required but not declared. The skill does not request unrelated cloud credentials, but it does rely on a privileged local session (Instagram) that enables posting on the user's behalf.
Persistence & Privilege
The skill does not request always:true or attempt to modify other skills or system-wide agent settings. It runs as a normal skill and uses OpenClaw's browser automation and local temp directories (/tmp/openclaw/uploads). Autonomous invocation is allowed by default but not unusual. No evidence of persistent background services or self-enabling behavior.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install veezvg-episode-to-instagram - After installation, invoke the skill by name or use
/veezvg-episode-to-instagram - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
- Adds end-to-end pipeline for converting podcast/video episodes into Instagram-ready content (carousel posts, quote cards, and Reels).
- Automates transcription, content extraction, video frame selection, image generation, and posting via browser automation.
- Supports both YouTube URLs and direct video files as input sources.
- Integrates explicit visual and caption approval steps before publishing.
- Includes strict content and branding rules, with customizable options in brand-config.json.
- Ensures posts are only uploaded after final user confirmation with visual previews.
Metadata
Frequently Asked Questions
What is Openclaw Skills?
Full pipeline to turn a video podcast episode into Instagram content (carousel posts, quote cards, and Reels). Transcribes the episode, extracts key quotes w... It is an AI Agent Skill for Claude Code / OpenClaw, with 112 downloads so far.
How do I install Openclaw Skills?
Run "/install veezvg-episode-to-instagram" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Openclaw Skills free?
Yes, Openclaw Skills is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Openclaw Skills support?
Openclaw Skills is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Openclaw Skills?
It is built and maintained by veezVg (@veezvg); the current version is v1.0.1.
More Skills