Material Report

Analyze ad material videos and produce a markdown report with framework, material traits, and acquisition keywords, then propose new material production fram...

Security Scan

The skill's requirements and runtime instructions align with its stated purpose: it analyzes a provided video, may read/write frame images locally, and produces a markdown report using the included template; it does not request credentials or external network access.

Purpose & Capability

Name/description (analyze ad videos and produce a report) match the instructions: verifying a video, extracting frames, analyzing structure, and producing a markdown report. No unrelated credentials, binaries, or services are requested.

Instruction Scope

Instructions legitimately require filesystem access to the provided video and may write extracted frame images into a folder under the current working directory. This is expected for video analysis but worth noting: the skill will check paths, read metadata, and may save many image files. It does not instruct the agent to read unrelated files, access environment variables, or transmit data externally.

Install Mechanism

No install spec or remote downloads — the skill is instruction-only and relies on ffmpeg being present on the user's system. The skill explicitly prompts the user to install ffmpeg if missing (but does not provide installation commands). This is low-risk.

Credentials

No environment variables, credentials, or config paths are requested. All required inputs are user-supplied (video path, category/vertical, platform).

Persistence & Privilege

Skill is not marked always:true and does not request persistent or cross-skill configuration. Autonomous invocation is allowed by default but is not combined with other concerning privileges.

Assessment

This skill appears coherent and matches its description. Before installing or running it, be aware that: it will need access to the video file you provide and may create a directory of extracted frame images in the current working directory (potentially many files). If you prefer not to have files written, provide an existing frame-folder path or run the skill from a controlled directory. The skill does not request credentials or network endpoints, but avoid supplying sensitive videos you do not want processed locally. If you want more assurance, you can ask for a dry-run that only reads metadata (no frame extraction) or run the skill in an isolated workspace.

Quick Start

• Request or confirm: video path, category/vertical, target platform.
• Duration spec: keep the new material duration roughly consistent with the original video unless the user explicitly requests otherwise.
• Analyze the provided video and output a markdown report.

Workflow

1. Verify the video file exists and read basic metadata (duration, resolution, fps).
2. Before extracting frames, check whether frame images already exist. If they do, reuse them directly.
3. If frame extraction is needed, use ffmpeg to extract frames. Save all frame images in a folder named after the video (same base name) under the current working directory. If permissions prevent this, ask the user to grant access; do not change the output path.
4. If ffmpeg is not available on the user's machine, prompt the user to install ffmpeg (do not provide installation instructions in this skill) or provide an existing frame image folder path.
5. Extract the original framework as ordered stages (opening hook, problem, turning points, resolution, ending/CTA).
   • Hook identification: prioritize the first 0-5 seconds and any large on-screen headline, abrupt visual change, or explicit "attention" phrasing. If the hook is ambiguous, state multiple candidates and pick the strongest with a brief rationale.
6. Summarize material traits (format, tempo, overlays, UI density, meme usage, etc.).
7. List acquisition keywords (concise, actionable, aligned to content and visual tactics).
8. Create at least one new material framework. For the "different direction" variant, choose the best-performing direction based on your judgment, without being constrained by the original material. Do not deviate from the same category/vertical and target platform.
9. Provide a detailed storyboard table for the new framework with time ranges, on-screen action, copy, and visual effects.
10. Deliver the report in markdown in the response.
11. If the user does not specify category/vertical, first ask for it; then proceed with a best-effort inference and label it clearly as an assumption in the report.

Output Template (Markdown)

• Load references/report_template.md and follow its structure

Notes

• Match the user's language and tone.
• Keep the storyboard single-level lists and clear time ranges.
• If the user requests a same-framework version, keep the structure identical and only change content details.