SkillsBench is a benchmark for measuring how effectively AI agents use modular skills—folders containing instructions, scripts, and resources—to complete specialized tasks. It helps researchers and developers evaluate both skill quality and agent behavior, including tasks that require combining multiple skills. The catalogue’s skills and instructions are evaluated as part of this workflow.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add benchflow-ai/skillsbench --skill ffmpeg-video-editinggit clone --depth 1 https://github.com/benchflow-ai/skillsbenchWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/benchflow-ai/skillsbench/ffmpeg-video-editing)<a href="https://agentmods.dev/skills/benchflow-ai/skillsbench/ffmpeg-video-editing"><img src="https://agentmods.dev/badge/skills/benchflow-ai/skillsbench/ffmpeg-video-editing/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/benchflow-ai/skillsbench/ffmpeg-video-editing"><img src="https://agentmods.dev/badge/skills/benchflow-ai/skillsbench/ffmpeg-video-editing.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00059 | $0.01039 |
| Opus 5 | $0.00030 | $0.00519 |
| Sonnet 5 | $0.00012 | $0.00208 |
| Haiku 4.5 | $0.00006 | $0.00104 |
Grade A, and why
ffmpeg-video-editing scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
result = subprocess.run([ How it starts
The opening of the file, as written. The whole thing — 141 lines — stays where its author put it; the contents beside it link to each section on GitHub.
FFmpeg Video Editing
Cutting Video Segments
Extract a portion (keep segment)
# Extract from start_time to end_time
ffmpeg -i input.mp4 -ss START -to END -c copy output.mp4
# With re-encoding for frame-accurate cuts
ffmpeg -i input.mp4 -ss START -to END -c:v libx264 -c:a aac output.mp4
Remove a segment (cut out middle)
To remove a segment, split into parts and concatenate:
# 1. Extract before the cut
ffmpeg -i input.mp4 -to CUT_START -c copy part1.mp4
# 2. Extract after the cut
ffmpeg -i input.mp4 -ss CUT_END -c copy part2.mp4
# 3. Concatenate
ffmpeg -f concat -safe 0 -i filelist.txt -c copy output.mp4
Concatenating Multiple Segments
Using concat demuxer (recommended for same-codec files)
Create a file list (segments.txt):
file 'segment1.mp4'
file 'segment2.mp4'
file 'segment3.mp4'
Then concatenate:
ffmpeg -f concat -safe 0 -i segments.txt -c copy output.mp4
Using filter_complex (for re-encoding)
ffmpeg -i seg1.mp4 -i seg2.mp4 -i seg3.mp4 \
-filter_complex "[0:v][0:a][1:v][1:a][2:v][2:a]concat=n=3:v=1:a=1[v][a]" \
-map "[v]" -map "[a]" output.mp4
Removing Multiple Segments (Batch)
For removing many short segments (like filler words), the efficient approach:
- Calculate the "keep" segments (inverse of remove segments)
- Extract each keep segment
- Concatenate all keep segments
import subprocess
import os
def remove_segments(input_file, segments_to_remove, output_file):
"""
segments_to_remove: list of (start, end) tuples in seconds
"""
# Get video duration
result = subprocess.run([
'ffprobe', '-v', 'error', '-show_entries', 'format=duration',
'-of', 'default=noprint_wrappers=1:nokey=1', input_file
], capture_output=True, text=True)
duration = float(result.stdout.strip())
# Sort segments and merge overlapping
segments = sorted(segments_to_remove)
# Calculate keep segments (gaps between remove segments)
keep_segments = []
current_pos = 0.0
for start, end in segments:
if start > current_pos:
keep_segments.append((current_pos, start))
current_pos = max(current_pos, end)
if current_pos < duration:
keep_segments.append((current_pos, duration))
# Extract each keep segment
temp_files = []
for i, (start, end) in enumerate(keep_segments):
temp_file = f'/tmp/seg_{i:04d}.mp4'
subprocess.run([
'ffmpeg', '-y', '-i', input_file,
'-ss', str(start), '-to', str(end),
'-c', 'copy', temp_file
], check=True)
temp_files.append(temp_file)
# Create concat list
list_file = '/tmp/concat_list.txt'
with open(list_file, 'w') as f:
for temp_file in temp_files:
f.write(f"file '{temp_file}'\n")
# Concatenate
subprocess.run([
'ffmpeg', '-y', '-f', 'concat', '-safe', '0',
'-i', list_file, '-c', 'copy', output_file
], check=True)
# Cleanup
for f in temp_files:
os.remove(f)
os.remove(list_file)
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 141 lines · 59 tokens per session scan A 4e33abec2186
ffmpeg-video-editing is a skill published in the GitHub repository benchflow-ai/skillsbench (1,754 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 59 tokens to every session and 1,039 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
opencv-template-matching
Use OpenCV for image processing, grayscale conversion, and template matching to detect objects in images.
exploded-view-illustration
Techniques for drawing isometric or orthographic exploded-view hardware diagrams programmatically using Pillow, with annotation leader lines and layer separation.
python-pillow-graphics
Create technical graphics, diagrams, and posters using Python Pillow library with precise color control and typography.
chinese-classical-poetry
A guide to writing a seven-character regulated poem in Classical Chinese, a traditional form with eight lines of seven Chinese characters each. It explains the required couplets, tone patterns, and rhymes.
image-generation-pillow
Technical image generation using Python's Pillow library. Use this when you need to programmatically create diagrams, posters, or technical drawings.
pillow-technical-drawing
Creating technical diagrams and exploded-view illustrations using Python Pillow with geometric primitives and text annotations.