一键在 Manus 中运行任何 Skill

agilab-product-reels

星标17

分支1

更新时间2026年6月3日 12:57

Build and refine short AGILAB product reels and technical demo videos with one-app storytelling, semantic guardrails, frame review, and YouTube packaging.

安装

用 Codex 或 Claude 帮你安装复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。

在 Manus 中运行

来源

ThalesGroup

ThalesGroup/agilab

打开 GitHub 仓库查看创作者相关仓库

下载

在 Manus 中运行

Product Reels Skill (AGILAB)

Use this skill when the user wants a short AGILAB teaser, technical benchmark reel, live-demo plan, or YouTube-ready demo package.

Pick The Right Demo Type

Product teaser: product-first, fast, broad, one core claim.
Technical benchmark reel: same scenario, same load/seed, one meaningful variable changes.
Live walkthrough: real cursor capture for the final “hero” demo; use only after the story is already clear.

Do not mix multiple apps in one short demo.

Default App Positioning

flight_project: safest product teaser.
uav_queue_project: best short technical benchmark reel.

If the user wants a short public video, prefer one of those two first.

Core Story Rule

Every short AGILAB demo must have one central claim, then map that claim through:

PROJECT
ORCHESTRATE
WORKFLOW
ANALYSIS

Do not make a feature tour. A short reel needs one story only.

Semantic Guardrails

Do not show an empty replayable stage or empty generated snippet.
Do not end on decorative ANALYSIS; show real evidence.
Do not show a “map” without credible geographic meaning.
Do not imply unsupported product behavior.
When saying scalability, state whether it means worker distribution or data-volume / big-data scale.
If the app supports a real benchmark comparison, show that comparison instead of generic workflow text.

If the content feels weak, improve the story before polishing the visuals.

Reel Workflow

Choose one app.
Write one sentence for the core claim.
Decide which content must be visible in each page:
- PROJECT: scenario or app context
- ORCHESTRATE: runnable packaging / orchestration intent
- WORKFLOW: replayable snippet, saved stage content, or workflow graph
- ANALYSIS: visible outcome or comparison evidence
Generate or refresh the reel.
Extract fresh frames and inspect them before treating the reel as ready.
Only after the content is clear, prepare the YouTube title/description/comment package.

Current Generator

Generator: tools/build_product_demo_reel.py
Outputs: artifacts/demo_media/<variant>/

Typical runs:

uv run python tools/build_product_demo_reel.py --variant flight
uv run python tools/build_product_demo_reel.py --variant uav_queue

Known Good Narrative Patterns

`flight_project`

Tell the product story.
Keep the sequence simple and clean.
For scale wording, prefer data-volume / big-data tractability over worker-count claims unless the user explicitly wants cluster scale.

`uav_queue_project`

Tell a benchmark story, not a workflow story.
Keep the same scenario, load, and seed fixed.
Change one meaningful variable only, such as routing policy.
Show concrete evidence in ANALYSIS: drops, delay, bottleneck, route usage, queue buildup.

Review Loop

After every meaningful change, extract and inspect frames for at least:

PROJECT
ORCHESTRATE
WORKFLOW
ANALYSIS

Useful checks:

uv run python -m py_compile tools/build_product_demo_reel.py
/opt/homebrew/bin/ffmpeg -y -ss 00:00:03 -i artifacts/demo_media/flight/agilab_flight.mp4 -frames:v 1 /tmp/frame.png
ffprobe -v error -select_streams v:0 -show_entries stream=width,height,r_frame_rate,bit_rate -of default=noprint_wrappers=1 artifacts/demo_media/flight/agilab_flight.mp4

Do not trust the code alone; verify the rendered frame.

YouTube Packaging

Once the reel is visually and semantically stable, prepare:

title
description
chapters
pinned comment
thumbnail text

Keep the upload copy aligned with the reel type:

teaser: product-first
benchmark reel: scenario-first and evidence-first
live walkthrough: action-first and page-by-page

Escalation Rule

If repeated iterations still look synthetic or “not pro”, stop tuning the composed reel and switch to a real live capture plan. Use the reel as intro/outro material only.

同仓库更多 Skills

同仓库

repo-skill-maintenance

ThalesGroup/agilab

Maintain repo-managed agent skills across `.claude/skills` and `.codex/skills`, including targeted sync, validation, index regeneration, drift checks, and Tokki skill visibility. Use when adding or updating a shared skill, migrating a user-managed skill into the repo, or reconciling agent skill copies without overwriting unrelated skills.

2026-07-1717

agilab-security-review-patterns

ThalesGroup/agilab

Review AGILAB changes for security hardening risks. Use when code, docs, or workflows touch installers, Streamlit exposure, cluster/SSH/share behavior, app execution, notebooks, LLM connectors, secrets, PyPI/GitHub/Hugging Face publishing, dependency policy, or external repositories.

2026-07-1617

agilab-intent-router

ThalesGroup/agilab

Route terse AGILAB operator requests such as "do it", "review AGILAB", "next move", "update repos", "merge it", "check again", "release", and "cluster validation" into the right repo skills, safety mode, validation depth, and output contract using session-derived policy.

2026-07-0717

agilab-streamlit-pages

ThalesGroup/agilab

Streamlit page authoring patterns for AGILAB (session_state safety, keys, rerun, UX).

2026-07-0417

slides

ThalesGroup/agilab

Create, edit, inspect, or explain presentation slide decks (`.pptx`) with PptxGenJS, bundled layout helpers, and render/validation utilities. Use when tasks involve building a new PowerPoint deck, recreating slides from screenshots/PDFs/reference decks, modifying content while preserving editable output, extracting or explaining a specific slide, adding charts/diagrams/visuals, or diagnosing layout issues such as overflow, overlaps, and font substitution.

2026-07-0117

agilab-deep-audit

ThalesGroup/agilab

Produce deep AGILAB audit and code-review artifacts with evidence-backed findings, mandatory architecture-foundation readiness, blast-radius tracing, security/test posture, and prioritized recommendations. Use when the user says "review AGILAB", "audit AGILAB", "code review AGILAB", "deep review", "architecture review", "security review", asks for a review document, or asks for comparison-quality critique rather than a quick fix.

2026-06-2417

name	agilab-product-reels
description	Build and refine short AGILAB product reels and technical demo videos with one-app storytelling, semantic guardrails, frame review, and YouTube packaging.
license	BSD-3-Clause (see repo LICENSE)
metadata	{"updated":"2026-05-07T00:00:00.000Z"}

Product Reels Skill (AGILAB)

Use this skill when the user wants a short AGILAB teaser, technical benchmark reel, live-demo plan, or YouTube-ready demo package.

Pick The Right Demo Type

Product teaser: product-first, fast, broad, one core claim.
Technical benchmark reel: same scenario, same load/seed, one meaningful variable changes.
Live walkthrough: real cursor capture for the final “hero” demo; use only after the story is already clear.

Do not mix multiple apps in one short demo.

Default App Positioning

flight_project: safest product teaser.
uav_queue_project: best short technical benchmark reel.

If the user wants a short public video, prefer one of those two first.

Core Story Rule

Every short AGILAB demo must have one central claim, then map that claim through:

PROJECT
ORCHESTRATE
WORKFLOW
ANALYSIS

Do not make a feature tour. A short reel needs one story only.

Semantic Guardrails

Do not show an empty replayable stage or empty generated snippet.
Do not end on decorative ANALYSIS; show real evidence.
Do not show a “map” without credible geographic meaning.
Do not imply unsupported product behavior.
When saying scalability, state whether it means worker distribution or data-volume / big-data scale.
If the app supports a real benchmark comparison, show that comparison instead of generic workflow text.

If the content feels weak, improve the story before polishing the visuals.

Reel Workflow

Choose one app.
Write one sentence for the core claim.
Decide which content must be visible in each page:
- PROJECT: scenario or app context
- ORCHESTRATE: runnable packaging / orchestration intent
- WORKFLOW: replayable snippet, saved stage content, or workflow graph
- ANALYSIS: visible outcome or comparison evidence
Generate or refresh the reel.
Extract fresh frames and inspect them before treating the reel as ready.
Only after the content is clear, prepare the YouTube title/description/comment package.

Current Generator

Generator: tools/build_product_demo_reel.py
Outputs: artifacts/demo_media/<variant>/

Typical runs:

uv run python tools/build_product_demo_reel.py --variant flight
uv run python tools/build_product_demo_reel.py --variant uav_queue

Known Good Narrative Patterns

`flight_project`

Tell the product story.
Keep the sequence simple and clean.
For scale wording, prefer data-volume / big-data tractability over worker-count claims unless the user explicitly wants cluster scale.

`uav_queue_project`

Tell a benchmark story, not a workflow story.
Keep the same scenario, load, and seed fixed.
Change one meaningful variable only, such as routing policy.
Show concrete evidence in ANALYSIS: drops, delay, bottleneck, route usage, queue buildup.

Review Loop

After every meaningful change, extract and inspect frames for at least:

PROJECT
ORCHESTRATE
WORKFLOW
ANALYSIS

Useful checks:

uv run python -m py_compile tools/build_product_demo_reel.py
/opt/homebrew/bin/ffmpeg -y -ss 00:00:03 -i artifacts/demo_media/flight/agilab_flight.mp4 -frames:v 1 /tmp/frame.png
ffprobe -v error -select_streams v:0 -show_entries stream=width,height,r_frame_rate,bit_rate -of default=noprint_wrappers=1 artifacts/demo_media/flight/agilab_flight.mp4

Do not trust the code alone; verify the rendered frame.

YouTube Packaging

Once the reel is visually and semantically stable, prepare:

title
description
chapters
pinned comment
thumbnail text

Keep the upload copy aligned with the reel type:

teaser: product-first
benchmark reel: scenario-first and evidence-first
live walkthrough: action-first and page-by-page

Escalation Rule

If repeated iterations still look synthetic or “not pro”, stop tuning the composed reel and switch to a real live capture plan. Use the reel as intro/outro material only.