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。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/diegosouzapw/awesome-omni-skill --skill ffmpeg-media命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Token-efficient tracking for AI orchestration. CLI-first for status updates (~50 tokens), agent fallback for complex ops (~1KB). Use when: updating task status, querying blockers, creating progress files, validating phases.
AshAi extension guidelines for integrating AI capabilities with Ash Framework. Use when implementing vectorization/embeddings, exposing Ash actions as LLM tools, creating prompt-backed actions, or setting up MCP servers. Covers semantic search, LangChain integration, and structured outputs.
This skill should be used when solving hard questions, complex architectural problems, or debugging issues that benefit from GPT-5 Pro or GPT-5.1 thinking models with large file context. Use when standard Claude analysis needs deeper reasoning or extended context windows.
基于 SOC 职业分类
正在显示 SKILL.md
| name | ffmpeg-media |
| description | FFmpeg media processing. Video/audio transcoding, stream manipulation, and filter graphs. |
| version | 1.0.0 |
Trit: +1 (PLUS - generative media transformation)
Foundation: FFmpeg + libav + filter system
FFmpeg transforms media through:
# Transcode video
ffmpeg -i input.mp4 -c:v libx264 -crf 23 output.mp4
# Extract audio
ffmpeg -i video.mp4 -vn -c:a aac output.m4a
# Convert to GIF
ffmpeg -i input.mp4 -vf "fps=10,scale=320:-1" output.gif
# Cut segment
ffmpeg -i input.mp4 -ss 00:01:00 -t 00:00:30 -c copy segment.mp4
# Concat files
ffmpeg -f concat -i list.txt -c copy output.mp4
# Scale and add text
ffmpeg -i input.mp4 \
-vf "scale=1280:720,drawtext=text='Title':fontsize=24:x=10:y=10" \
output.mp4
# Color adjustment
ffmpeg -i input.mp4 -vf "eq=brightness=0.1:saturation=1.2" output.mp4
def trit_from_media_op(op: str) -> int:
"""Map FFmpeg operations to GF(3) trits."""
if op in ["probe", "analyze", "check"]:
return -1 # MINUS: verification
elif op in ["copy", "remux", "extract"]:
return 0 # ERGODIC: preservation
else:
return 1 # PLUS: transformation
spi-parallel-verify (-1) ⊗ video-downloader (0) ⊗ ffmpeg-media (+1) = 0 ✓
mathpix-ocr (-1) ⊗ image-enhancer (0) ⊗ ffmpeg-media (+1) = 0 ✓
This skill connects to the K-Dense-AI/claude-scientific-skills ecosystem:
general: 734 citations in bib.duckdbThis skill connects to Software Design for Flexibility (Hanson & Sussman, 2021):
Concepts: combinators, compose, parallel-combine, spread-combine, arity
ffmpeg-media (○) + SDF.Ch1 (+) + [balancer] (−) = 0
Skill Trit: 0 (ERGODIC - coordination)
Combinators compose operations. This skill provides composable abstractions.
This skill maps to Cat# = Comod(P) as a bicomodule in the equipment structure:
Trit: 0 (ERGODIC)
Home: Prof
Poly Op: ⊗
Kan Role: Adj
Color: #26D826
The skill participates in triads satisfying:
(-1) + (0) + (+1) ≡ 0 (mod 3)
This ensures compositional coherence in the Cat# equipment structure.