agentleFS
Eugenevdm01Eugenevdm01/eugenes-meta-agent/skills/funnel-junkies-video-publisher/SKILL.md

funnel-junkies-video-publisher

Skill4 starsChanged 6 months agoMITView on GitHub

Funnel Junkies short-form video editing, captioning, approval, and VanX/GHL Social Planner publishing workflow. Use for: processing raw phone or mic-recorded videos into captioned Reels, preserving original mic audio, preparing approval packages, scheduling approved posts to VanX-connected Facebook/Instagram channels, and logging publishing confirmations.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill when Eugene asks to edit, caption, package, approve, post, schedule, or test a short-form Funnel Junkies video for organic …

    line 8
  2. Distinctive:

    Wait for explicit approval: before scheduling or publishing.

    line 26
  3. Distinctive:

    Schedule the live Reel: only after approval. Use at least a 60-minute future buffer unless immediate publishing has been proven reliable …

    line 30

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Funnel Junkies Video Publisher

Use this skill when Eugene asks to edit, caption, package, approve, post, schedule, or test a short-form Funnel Junkies video for organic social publishing through VanX / GoHighLevel Social Planner.

Non-Negotiable Rule

Preserve the original recorded mic audio for all future videos. Until Eugene separately approves a fixed sound-editing process, do not enhance, clean, normalize, compress, denoise, EQ, level, replace, re-encode, or otherwise alter the audio under any circumstances.

If the output container cannot copy the original audio stream, stop and ask before re-encoding. Treat audio changes as blocked by default, not as an optional editing step. Even if the video sounds imperfect, keep the original sound unless Eugene explicitly approves a dedicated sound-editor test or revised audio workflow.

Core Workflow

  1. Create a job folder under /home/ubuntu/ or the persistent video agent workspace. Store the raw source, transcript, captions, renders, handoff metadata, approval notes, and VanX result files in one place.
  2. Probe the source video with ffprobe. Record duration, video dimensions, orientation, frame rate, audio codec, audio sample rate, and stream mapping.
  3. Transcribe the source audio with manus-speech-to-text or an equivalent transcription tool. Save the timed transcript JSON.
  4. Generate captions with scripts/create_ass_captions.py. Review the transcript for obvious recognition errors before final export.
  5. Render the captioned Reel with scripts/render_captioned_reel_copy_audio.py. This burns in captions, formats video for portrait social delivery, and uses -c:a copy to preserve the original mic audio.
  6. Verify the audio with scripts/verify_audio_stream.py. Save the JSON report and do not proceed if the source and edited audio hashes do not match, unless Eugene explicitly approved audio re-encoding.
  7. Create handoff metadata using references/handoff_schema.md. Include caption, hashtags, final video path, public video URL if available, target channels, audio policy, and approval status.
  8. Create an approval package using templates/approval_package_template.md. Attach the final video and approval package to Eugene. Clearly state that nothing has been published yet.
  9. Wait for explicit approval before scheduling or publishing. If Eugene says “post it”, “publish”, “approve”, or equivalent, proceed to the publishing phase.
  10. Upload the approved MP4 with manus-upload-file so VanX can ingest a public MP4 URL.
  11. Before using VanX, read /home/ubuntu/skills/vanx-social-planner/SKILL.md if it exists, then use the existing VanX posting helpers or scripts/publish_vanx_reel.py.
  12. Run a dry-run first to confirm the exact target channels and payload. Do not live-submit to channels that were not requested or present in the approved handoff.
  13. Schedule the live Reel only after approval. Use at least a 60-minute future buffer unless immediate publishing has been proven reliable for that VanX account.
  14. Save live result files for each platform and create publishing_confirmation.md with scheduled time, channel, platform, post IDs, status, and the audio-preservation note.
  15. Report completion to Eugene with the publishing confirmation, final MP4, handoff JSON, and live result JSON files attached.

Bundled Resources

ResourcePurpose
scripts/create_ass_captions.pyConvert timed transcription JSON into styled ASS captions and a clean text transcript.
scripts/render_captioned_reel_copy_audio.pyRender a 1080x1920 captioned Reel while copying the original source audio stream unchanged.
scripts/verify_audio_stream.pyCompare source and final audio MD5 hashes to prove audio preservation.
scripts/publish_vanx_reel.pyDry-run or schedule approved Reels through VanX/GHL Social Planner using a public video URL.
references/workflow_checklist.mdDetailed operational checklist and quality gates.
references/handoff_schema.mdStandard organic_agent_handoff.json structure.
templates/approval_package_template.mdReusable pre-publication approval package template.

Standard Commands

Generate captions:

python /home/ubuntu/skills/funnel-junkies-video-publisher/scripts/create_ass_captions.py \
  --transcript-json transcription.json \
  --ass-out captions.ass \
  --text-out transcript_clean.txt

Render with original audio copied:

python /home/ubuntu/skills/funnel-junkies-video-publisher/scripts/render_captioned_reel_copy_audio.py \
  --source source.mp4 \
  --captions captions.ass \
  --out final_original_audio.mp4 \
  --report render_report.json

Verify audio preservation:

python /home/ubuntu/skills/funnel-junkies-video-publisher/scripts/verify_audio_stream.py \
  --source source.mp4 \
  --edited final_original_audio.mp4 \
  --out audio_verification.json

Dry-run VanX scheduling:

python /home/ubuntu/skills/funnel-junkies-video-publisher/scripts/publish_vanx_reel.py \
  --video-url "https://public-file-url/final_original_audio.mp4" \
  --handoff-json organic_agent_handoff.json \
  --target "Work from NoWhere - Funnel Junkies" \
  --target "ecomm_adventures" \
  --minutes-ahead 60 \
  --out vanx_dry_run_result.json

Live VanX scheduling after Eugene approves:

python /home/ubuntu/skills/funnel-junkies-video-publisher/scripts/publish_vanx_reel.py \
  --video-url "https://public-file-url/final_original_audio.mp4" \
  --handoff-json organic_agent_handoff.json \
  --target "Work from NoWhere - Funnel Junkies" \
  --target "ecomm_adventures" \
  --minutes-ahead 60 \
  --live \
  --out vanx_live_result.json

Quality Gates

GateRequired Standard
AudioSource and final audio hashes must match. Audio changes remain blocked for all future videos until Eugene approves a fixed sound-editing workflow.
CaptionsCaptions must be readable on mobile and positioned away from critical face/CTA areas.
ApprovalNo live publishing without explicit approval from Eugene after the final asset is attached.
TargetingDry-run must confirm the exact intended VanX channels before live submission.
ConfirmationFinal response must include status, scheduled time, platform, channel, and result files.

Default Tested Targets

The tested workflow successfully scheduled a corrected Reel to Work from NoWhere - Funnel Junkies as a Facebook Reel and ecomm_adventures as an Instagram Reel. Confirm targets each time because connected VanX accounts may change.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.