by xwchris
# Add to your Claude Code skills
git clone https://github.com/xwchris/videosays-agent-toolsGuides for using ai agents skills like videosays-agent-tools.
videosays-agent-tools is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by xwchris. It has 0 GitHub stars.
videosays-agent-tools's catalog security scan is still queued. You can run an instant dependency and prompt-injection check now with the "Scan for vulnerabilities" button above.
Clone the repository with "git clone https://github.com/xwchris/videosays-agent-tools" and add it to your Claude Code skills directory (see the Installation section above). videosays-agent-tools ships a SKILL.md manifest, so compatible agents can discover and load it automatically.
videosays-agent-tools is primarily written in JavaScript. It is open-source under xwchris on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh videosays-agent-tools against similar tools.
No comments yet. Be the first to share your thoughts!
Unlocks once the catalog security scan passes (runs nightly).
The deep catalog scan for this skill is still queued. Run an instant dependency check now instead.
name: videosays version: 1.0.3 description: Videosays video transcription, video to text, speech to text, subtitle extractor, caption transcription, YouTube transcript, TikTok transcript, Douyin transcript, Xiaohongshu transcript, and AI agent video transcription. Use when the user asks to transcribe a video link, extract spoken text, generate subtitles, check credit balance, or view transcription history. license: MIT-0 requires: binaries: - npx sendsDataTo:
Use npx videosays to turn video links or share text into transcript text. Search terms this skill covers include video to text, video transcription, speech to text, subtitle extraction, caption transcription, Douyin transcription, TikTok transcription, YouTube transcription, Xiaohongshu transcription, and short-video transcript.
This skill sends the user's API key and submitted video link/share text to Videosays.
npxIf no API key is configured, run:
npx videosays login
The CLI prints a browser authorization URL and waits. Ask the user to open that URL, sign in to Videosays, and authorize the CLI. The CLI then saves the API key to ~/.videosays.
If the user provides an API key directly, run:
npx videosays login --api-key "$VIDEOSAYS_API_KEY"
Do not print or reveal the API key in responses.
Use the simplest command that matches the user's requested output. By default the CLI prints only transcript text to stdout, which is the preferred agent workflow.
# Check whether the CLI is already authenticated
npx videosays whoami
# Transcribe a video link or share text
npx videosays transcribe "https://www.tiktok.com/@creator/video/123456"
# Explicitly request plain text, the default output
npx videosays transcribe "https://v.douyin.com/xxxxx/" --format text
# Include timestamps for each segment
npx videosays transcribe "https://www.tiktok.com/@creator/video/123456" --format timeline
# Generate subtitle formats
npx videosays transcribe "https://www.tiktok.com/@creator/video/123456" --format srt
npx videosays transcribe "https://www.tiktok.com/@creator/video/123456" --format vtt
# Check a task that is still running
npx videosays status "<task-id>"
npx videosays status "<task-id>" --format srt
# Query credits
npx videosays balance
# View recent transcription history
npx videosays history
Before the first transcription in a session, run:
npx videosays whoami
If it exits successfully, continue with transcription. If it exits non-zero because no API key is configured, run:
npx videosays login
Ask the user to complete the browser authorization URL printed by the CLI, then retry npx videosays whoami. Do not ask the user for their API key unless they explicitly want to provide one.
Long videos may still be processing when transcribe returns. If stdout contains this block:
VIDEOSAYS_TASK_PENDING
task_id=<task-id>
status=<status>
next=videosays status <task-id>
Do not treat it as a final transcript. Extract task_id, wait a reasonable interval, then run the next command. If the original user requested a format, preserve it on status checks, for example:
npx videosays status "<task-id>" --format srt
Repeat until the command prints transcript/subtitle content or exits with an error.
If a command exits non-zero, read stderr. Do not treat stdout as transcript content.
The CLI prints stable error details when available:
Error: <message>
Code: <error-code>
Next: <recommended-command>
Recharge: <billing-url>
If Code: insufficient_credits appears, do not retry transcription. Tell the user their balance is insufficient, suggest npx videosays balance, and provide the recharge URL if present.
For media or link errors such as media_resolve_failed, media_unavailable, or media_inaccessible, tell the user the link could not be processed and ask for another video link.
This skill calls npx videosays, which sends the submitted video link/share text and API key to https://api.videosays.com for transcription. Transcript text or the requested timestamp/subtitle format is returned on stdout.
Videosays command-line and agent skill packages for turning video links and share text into clean transcript text.
This repository contains:
packages/cli - npm CLI package, exposed as the videosays command.skills/videosays - canonical agent skill that calls npx videosays.SKILL.md - root copy for directories that require a repository-root skill file, such as SkillsLLM.npx videosays login
npx videosays transcribe "https://www.tiktok.com/@creator/video/123456"
The CLI stores your API key in ~/.videosays. You can also provide it through VIDEOSAYS_API_KEY.
Use Videosays from an AI agent by giving it this prompt:
Read https://videosays.com/SKILL.md and help me with Videosays.
Or install the skill from this repository:
npx skills add xwchris/videosays-agent-tools
skills/videosays/SKILL.md is the canonical skill entry for registry distribution. The root SKILL.md and the website copy at https://videosays.com/SKILL.md should stay byte-for-byte in sync with it.
skills.sh discovers skills from public GitHub repositories. This repository uses the standard skills/videosays/SKILL.md layout:
npx skills add xwchris/videosays-agent-tools
npx skills find videosays
Install the published skill from the Wegofuture organization:
clawhub install @wegofuture/videosays
clawhub inspect @wegofuture/videosays
Publish updates after signing in to ClawHub with access to the wegofuture publisher:
clawhub login
clawhub skill publish skills/videosays --owner wegofuture
SkillUse can install skills from GitHub-backed repositories:
skilluse login
skilluse repo add xwchris/videosays-agent-tools --path skills --branch main --default
skilluse install videosays
Keep the canonical skill, root skill copy, and website copy in sync:
bash scripts/check-skill-sync.sh
VIDEOSAYS_PUBLIC_SKILL_PATH=/path/to/video2txt/packages/web/public/SKILL.md bash scripts/check-skill-sync.sh
SkillsLLM expects a SKILL.md file at the root of the submitted GitHub repository. Submit this repository URL and use the root SKILL.md copy:
https://github.com/xwchris/videosays-agent-tools
videosays login
videosays login --api-key <api-key>
videosays logout
videosays whoami
videosays transcribe <video-link-or-share-text>
videosays transcribe <video-link-or-share-text> --format text
videosays transcribe <video-link-or-share-text> --format timeline
videosays transcribe <video-link-or-share-text> --format srt
videosays transcribe <video-link-or-share-text> --format vtt
videosays status <taskId>
videosays status <taskId> --format srt
videosays balance
videosays history [limit]
videosays help
videosaysMIT