ai-presenter-video skill

AI Presenter Video

Turn a topic (or finished script) plus ONE authorized adult presenter image into a complete, publish-ready presenter-led video: locked narration, avatar generation with lip-sync QA, captions, deterministic editing, loudness-normalized master/share encodes, and machine + visual acceptance reports.

Use this skill for new presenter videos AND for continuing, revising, captioning, lip-sync-repairing, or re-exporting an existing presenter-video job. The workflow is provider-neutral: pick generation capabilities from what is actually available in the session (FAL video/image models via imagegenerate and the video-gen plugin, TTS via textto_speech, ASR via the whisper/STT tooling, ffmpeg for everything deterministic).

Ported from cclank/lanshu-create-ai-presenter-video (MIT). Upstream body

kept substantively verbatim in references/; Hermes adaptations live in

this hub file. Scripts are deterministic (no network, no credentials).

Hermes adaptations (read first)

In Hermes the loader expands ${HERMESSKILLDIR} to this skill's installed directory, so every command below uses that token directly:

• Skill dir resolution — upstream hardcoded its own agent's skills path.

SKILL_DIR="${HERMES_SKILL_DIR}"

Shell variables do not persist between tool calls — re-paste the assignment (or the expanded path) in each terminal call that uses it.

capability", use texttospeech (OpenAI/Edge/ElevenLabs per user config); "presenter/avatar generation" → FAL image-to-video families (Kling, Wan, MiniMax H3 etc.) through the configured video tooling, or an avatar/lipsync endpoint the user has access to; "word-timestamp ASR" → whisper via the STT tooling or faster-whisper in a venv; "deterministic compositor" → ffmpeg filtergraphs, or the hyperframes skill when installed (the editing reference has a HyperFrames section that maps directly onto it).

• Capability mapping — where the references say "a voice generation

vision_analyze on the generated contact sheet plus sampled frames (identity, mouth timing, hands, blinking, continuity). Numeric checks come from the scripts' ffprobe output.

• Visual QA — do the "normal-speed visual review" steps with

Follow the upstream operating rules: before the first paid call state the uploaded assets, requested seconds, known cost, pilot size, and retry ceiling, and get the user's explicit go-ahead. Never upload the presenter image to a remote provider before remoteuploadapproved is true in job.json.

• Paid-generation consent — remote avatar/TTS generation is billable.

adultpresenterconfirmed, remoteuploadapproved, and voicecloneapproved sit inside the input object of job.json (init flags set them; hand-editing must target input., not the job root). manualinputreview. sits at the root. preflight.py distinguishes errors (block everything) from remote_blockers (block only remote generation) — local script/audio work may proceed while remote is blocked.

• Consent flags live under input — rights_confirmed,

Workflow

1. Start or resume a job. New job:

python3 "$SKILL_DIR/scripts/init_job.py" \
     --job-dir ~/Videos/my-presenter-video \
     --presenter-image /path/to/presenter.png \
     --topic "explain context engineering in one minute" \
     --duration 60 --aspect 9:16 \
     --rights-confirmed --adult-presenter-confirmed

Use --script for an existing script file; other flags: --voice-sample, --supporting-media, --width, --height, --fps, --watermark, --cta. For an existing job, read job.json + QA reports and resume from the earliest unfinished state — never regenerate accepted work.

(visionanalyze) and listen to any voice sample; record findings by setting the manualinput_review booleans in job.json, e.g.:

1. Manual input review. Actually look at the presenter image

python3 - <<'PY'
   import json
   p = "~/Videos/my-presenter-video/job.json"  # expand ~ or use an absolute path
   import os; p = os.path.expanduser(p)
   j = json.load(open(p))
   j["manual_input_review"].update(image_viewed=True, single_clear_face=True,
                                   image_has_no_unwanted_text=True)
   json.dump(j, open(p, "w"), indent=2)
   PY

Then gate:

python3 "$SKILL_DIR/scripts/preflight.py" ~/Videos/my-presenter-video/job.json

Proceed only when ok: true; do remote generation only when remote_ready: true. Note: preflight also updates job.json in place (records the report path) — re-read it after running rather than editing a stale copy.

full narration via texttospeech → ASR-verify the narration against the script → record real durations. The locked audio is the master clock for everything downstream.

1. Lock content and audio — read references/generation.md. Script →

Short low-cost pilot first; full run only after the pilot passes identity and mouth-timing review.

1. Plan and generate the presenter — read references/generation.md.

the locked audio; captions and keyword callouts only after audio and media are final.

1. Edit — read references/editing.md. Deterministic timeline driven by

1. Verify and deliver — read references/qa-recovery.md, render, then:

bash "$SKILL_DIR/scripts/finalize_delivery.sh" \
     ~/Videos/my-presenter-video/renders/rendered.mp4 \
     ~/Videos/my-presenter-video/outputs my-video

The finalizer preserves aspect ratio, runs two-pass loudness normalization (program ≈ −16 LUFS), produces master + share encodes, decode-verifies both, writes a delivery report JSON, and emits a nine-frame contact sheet. Inspect the contact sheet with vision_analyze before claiming completion.

Operating rules (non-negotiable)

voice-cloning authorization before the relevant remote action.

• Confirm image rights, adult status, remote-upload approval, and

sample or a stock TTS voice.

• Never infer or clone a real person's voice from an image; use an authorized

final scene boundaries.

• Lock the complete narration before presenter generation, caption timing, or

and intentional mix tracks carry audio.

• Mute video sources in the final composition; only the approved narration

URLs). Poll interrupted work before resubmitting — avoid double billing.

• Preserve provider request bodies and task IDs (minus credentials/expiring

sheet or full playback has been reviewed.

• Stop after three rejected paid candidates and summarize the failure mode.
• Do not claim completion until the final files fully decode and the contact

Defaults for minimal input

9:16, 1080×1920, 30fps; topic-derived videos target 45–75s; stock voice when no authorized sample; presenter-led layout with hook → 2–4 beats → close; no music/CTA unless requested; language inferred from the request.

Reference routing

presenter prompts, paid generation, provider changes.

• references/generation.md — intake, content, voice, capability selection,

keyword-callout presets, HyperFrames composition, exports.

• references/editing.md — timeline contract, openings/closes, captions,

recovery for lip-sync/identity/hands/exposure/freeze/caption/audio faults.

• references/qa-recovery.md — technical acceptance, visual acceptance, and