decompose-multimedia-source-verification
A two-stage strategy to isolate specific visual or bibliographic details from social media and video content by separating source identification from contextual extraction.
DeepseekModel
Curated skill
Quality Good · 64
v1.0.0
Get
https://deepseekmodel.com/api/download.php?id=web2bigtable-web2bigtable-orchestrator-skills-decompose-multimedia-source-verification-skill-md&format=skill
Download .skill
Standard format with system_prompt and model_config, ready for any agent framework
The actual content of the system_prompt field in the .skill file.
name decompose-multimedia-source-verification description A two-stage strategy to isolate specific visual or bibliographic details from social media and video content by separating source identification from contextual extraction. When to Use This strategy is triggered when a query requires identifying a specific object, title, or detail contained within a piece of multimedia content. It is applicable when the user provides metadata signals such as a platform-specific title, an uploader's handle, or a precise upload timeframe. Use this when the answer cannot be found in general knowledge but resides within the visual or auditory stream of a specific digital asset. Decomposition Template Source Authentication & Retrieval : Locate the primary source URL and retrieve the most granular metadata available (full transcript, OCR data, or frame-by-frame descriptions). The goal is to move beyond search engine snippets to the actual content body. Targeted Contextual Extraction : Within the retrieved source, isolate the specific segment referenced. Distinguish between the subject matter (what the video is about) and the target object (the specific item, text, or title requested). Cross-Reference Verification : Compare the extracted detail against the video’s official description or pinned comments to ensure the item is not a misidentified prop or a generic placeholder. Worker Assignment Rules Worker 1 (The Scout) : Responsible for finding the exact video link and the full transcript/metadata. Worker 2 (The Analyst) : Responsible for scanning the transcript or visual descriptions to find the specific timestamp and extracting the requested detail. Worker 3 (The Auditor) : Required if the target is a "title" or "name," to ensure the worker hasn't substituted a descriptive summary for a formal name. Answer Format The final output must clearly state the specific detail found, followed by the timestamp or source segment where it was located. 最终答案: [Specific Detail/Title] (Found at [Timestamp/Context]) Anti-Patterns Topic-Object Confusion : Substituting a general description of the video's theme for the specific bibliographic or physical title requested. Snippet Reliance : Formulating an answer based on search engine preview text which often truncates or summarizes the actual data. Temporal Drift : Failing to verify the upload date, leading to the selection of a similar video from the same creator but a different timeframe. Metadata Blindness : Ignoring the video description or pinned comments which often contain the precise names of items shown in the video.
Keywords that activate this skill. Click one to copy it.
This skill does not provide trigger words.
The downloaded .skill package contains the following fields.
| Field | Description |
|---|---|
| format | Format tag (skill/v1) |
| skill_id | Unique skill ID |
| name | Skill name |
| version | Version |
| description | Description |
| category | Categories (array) |
| trigger_words | Trigger words |
| tags | Tags |
| source | Source |
| source_url | Source URL (this page) |
| exported_at | Exported at (set per download) |
| system_prompt | System prompt body |
| model_config | Model config: provider / model / temperature / max_tokens / top_p |
| examples | Examples |
| install_guide | Import guide for Coze / Dify / Claude / custom frameworks |
The same skill can be exported in different platform formats.