Watchless

Watchless

English | 简体中文

Turn one video into a complete visual article. The transcript is the factual source; screenshots preserve the visual evidence. The reader should not need to watch the original video.

Hard Requirements

• Run locally. Do not call PodSum, its website, MCP, Cloudflare, APIFY, D1, or R2.
• Keep the source immutable and reuse cached downloads and transcripts.
• Use Tencent Cloud word-level ASR with speaker diarization by default (--provider tencent; auto also selects Tencent). Never automatically fall back to Volcengine; select other providers explicitly. Do not silently replace it with YouTube automatic captions or local Whisper.
• Keep the output chronological and complete. Light-plus is not a summary: preserve reasoning, examples, figures, caveats, disagreement, questions, answers, and repeated emphasis.
• Show process images during the run: whole-video route overview, boundary evidence where applicable, candidate frames, selected frames, and PDF page overview.
• Use no OpenCV anywhere in this Skill.
• Visual algorithms are route-specific. slides uses ffmpeg keyframe recall plus local SSIM refinement. explainer, conversation, and demo do not use SSIM, perceptual hash, histogram scoring, face scoring, or other visual ranking; Codex directly reads their candidate images.
• Resolve speaker identities for every conversation video before writing notes. Keep Speaker N when evidence remains weak or conflicting.
• Record model token usage and cost in work/token-usage.json whenever the runtime exposes real counts. Never invent unavailable usage or silently apply stale prices.
• Do not claim completion until HTML image references, rendered PDF pages, and ZIP contents pass verification.

Private-Use Compliance Gate

Before acquiring a remote source, confirm that the user owns the content, has permission to process it, or has independently established another lawful basis. If authorization is unclear, stop and request an authorized local file instead of downloading.

• This repository and its outputs are private by default. Do not publish or distribute generated notes without a separate rights, privacy, confidentiality, and attribution review.
• Chrome browser cookies may be read only from the user's local browser profile. Never export, print, copy, upload, log, or commit cookie values. Cookie access does not establish a right to download or reuse content.
• Never use this Skill to bypass DRM, paywalls, members-only access, private-video access, geographic restrictions, CAPTCHAs, account enforcement, or other access controls.
• Before sending non-public or sensitive audio to the selected cloud ASR provider, confirm that the user is authorized to make that third-party transfer. Use explicitly requested local Whisper or stop when authorization is unavailable.
• Resolve speaker names only from explicit public metadata, self-introduction, lower thirds, or similarly reliable evidence. Do not perform face or voice biometric identification. Keep Speaker N when uncertain and require human review before external use.
• Use the minimum screenshots and quotations needed for any approved external use. Do not publish a complete transcript or visual reconstruction that substitutes for the source without a separate legal review.

Read LEGAL.en.md before running this Skill on third-party, confidential, commercial, or sensitive material. These controls reduce risk but do not constitute legal advice.

Route Model

Choose one content route after directly inspecting verify/mode-overview.jpg and the transcript. Then add visual strategies. Do not route from the title or channel alone.

Visual strategies are composable:

• slide-state: stable, complete slide/build state.
• speaker: active speaker or useful group shot.
• evidence: chart, quote, interface, object, or source that carries the claim.
• broll: relevant external footage in an edited interview or documentary.
• dense-visual: increase candidate density for XiaoLin-style edited explainers.
• document-evidence: prefer papers, article excerpts, tables, and diagrams.
• screen-state: prefer a legible UI result rather than cursor motion or transitions.

For conversation, also choose one extraction profile without creating a new main route:

Typical routing examples:

• XiaoLin-style scripted finance/technology video: explainer + dense-visual + evidence.
• Best Partners paper/article explanation: explainer + document-evidence.
• a16z or YC Lightcone studio podcast: conversation + speaker.
• Silicon Valley 101 edited field interview: conversation + evidence + broll.
• YC Design Review screen walkthrough: demo + screen-state + evidence.
• Low-information entertainment chat: conversation + speaker, with fewer, larger topic units.

Workflow

Set SKILL_DIR to this Skill directory. Run commands from the workspace where outputs/video-notes/ should be created.

1. Acquire, transcribe, and inspect the whole video

"$SKILL_DIR/.venv/bin/python" "$SKILL_DIR/scripts/00_build_video_notes.py" "$SOURCE" --stage prepare

Read PROJECT_DIR from output. Display verify/mode-overview.jpg, inspect the transcript and source metadata, and decide the route from observable structure.

2. Confirm route and visual strategies

"$SKILL_DIR/.venv/bin/python" "$SKILL_DIR/scripts/00_build_video_notes.py" \
  --stage route --project-dir "$PROJECT_DIR" \
  --mode explainer \
  --visual-strategy dense-visual \
  --visual-strategy evidence \
  --mode-reason "scripted host alternates with explanatory charts and B-roll"

Later stages may use --mode auto; they read the confirmed work/route-decision.json. An explicit --mode remains a manual override. Compatibility aliases: presentation -> slides, editorial -> explainer.

For a Bloomberg-style news interview:

"$SKILL_DIR/.venv/bin/python" "$SKILL_DIR/scripts/00_build_video_notes.py" \
  --stage route --project-dir "$PROJECT_DIR" \
  --mode conversation --conversation-profile news \
  --visual-strategy speaker --visual-strategy evidence \
  --mode-reason "broadcast interview with lower thirds and sparse data graphics"