Global Rank · of 601 Skills
byted-tos-video-process AI Agent Skill
View Source: bytedance/agentkit-samples
SafeInstallation
npx skills add bytedance/agentkit-samples --skill byted-tos-video-process 10
Installs
Bytedance TOS Video Process Skill
This skill provides essential video processing functions for video files stored in Bytedance's TOS (TeraObjectStore). It allows you to retrieve video metadata and perform single-frame or multi-frame snapshots directly using the Volcengine TOS SDK.
Quick Start
1. Client Initialization
The following Python snippet demonstrates how to initialize the TosClientV2 from environment variables.
import os
import tos
from tos.exceptions import TosClientError, TosServerError
def create_client() -> tos.TosClientV2:
"""Initializes a TosClientV2 using AK/SK (and optional STS token) from environment variables."""
try:
ak = os.getenv('TOS_ACCESS_KEY')
sk = os.getenv('TOS_SECRET_KEY')
endpoint = os.getenv('TOS_ENDPOINT')
region = os.getenv('TOS_REGION')
security_token = os.getenv('TOS_SECURITY_TOKEN') # Optional, for STS
if not all([ak, sk, endpoint, region]):
raise ValueError("Required environment variables are missing (AK, SK, Endpoint, Region).")
return tos.TosClientV2(
ak=ak,
sk=sk,
endpoint=endpoint,
region=region,
security_token=security_token,
)
except (ValueError, ImportError) as e:
print(f"Error initializing client: {e}")
# Handle initialization failure
return None
# Create the client
client = create_client()2. Basic Workflow
# (Assumes 'client' is initialized and 'bucket_name', 'object_key' are set)
# 1. Get Video Info
try:
response = client.get_object(bucket_name, object_key, process="video/info")
info_data = response.read()
print("Video Info:", info_data.decode('utf-8'))
except TosServerError as e:
print(f"Error getting video info: {e}")
# 2. Take a Single Snapshot and save locally
try:
client.get_object_to_file(
bucket_name,
object_key,
"snapshot_1000ms.jpg",
process="video/snapshot,t_1000,f_jpg,w_720"
)
print("Snapshot saved to snapshot_1000ms.jpg")
except TosServerError as e:
print(f"Error taking snapshot: {e}")
# 3. Take a Snapshot and save back to TOS
try:
response = client.get_object(
bucket_name,
object_key,
process="video/snapshot,t_5000,f_jpg",
save_bucket=bucket_name,
save_object="processed/snapshot_5000ms.jpg"
)
save_result = response.read()
print("Snapshot saved to TOS:", save_result.decode('utf-8'))
except TosServerError as e:
print(f"Error saving snapshot to TOS: {e}")Core Operations
All video processing is achieved via the process parameter in the get_object or get_object_to_file SDK methods.
1. Get Video Info (videoInfo)
Retrieves metadata of a video file, such as resolution, duration, and format.
SDK Method: client.get_object(..., process="video/info")
# 'client', 'bucket_name', and 'object_key' must be defined
try:
response = client.get_object(bucket_name, object_key, process="video/info")
# The response body is a JSON string
video_metadata = response.read().decode('utf-8')
print(video_metadata)
except TosServerError as e:
print(f"Server Error: {e.code} - {e.message}")2. Take a Single Snapshot (videoSnapshot)
Captures a single frame from a video. It supports various parameters for customization and can either return the image data or save the result directly back to TOS.
SDK Method: client.get_object_to_file(..., process="video/snapshot,...") for local save.
SDK Method: client.get_object(..., process="video/snapshot,...", save_bucket=..., save_object=...) for saving to TOS.
# Example: Take a snapshot at 10 seconds, resize to 720p width, and save locally
try:
client.get_object_to_file(
bucket_name,
object_key,
file_path="local_snapshot.jpg",
process="video/snapshot,t_10000,w_720,f_jpg"
)
print("Snapshot saved successfully to local_snapshot.jpg")
except (TosClientError, TosServerError) as e:
print(f"An error occurred: {e}")3. Take Multiple Snapshots (videoSnapshots)
This is a client-side orchestration pattern. You loop through a series of timestamps and make multiple calls to the videoSnapshot operation. The scripts/video_snapshots.py provides a reference implementation for parallel execution.
# (Assumes 'client', 'bucket_name', 'object_key' are set)
timestamps = [1000, 5000, 10000] # In milliseconds
for i, ts in enumerate(timestamps):
output_filename = f'snapshot_{i+1}_at_{ts}ms.jpg'
process_rule = f"video/snapshot,t_{ts},w_720,f_jpg"
try:
client.get_object_to_file(
bucket_name,
object_key,
output_filename,
process=process_rule
)
print(f"Saved snapshot to {output_filename}")
except (TosClientError, TosServerError) as e:
print(f"Failed for timestamp {ts}: {e}")Authorization
Authentication is handled directly by the tos.TosClientV2 constructor. Provide credentials via environment variables.
Required Environment Variables
TOS_ACCESS_KEY: Your Access Key ID.TOS_SECRET_KEY: Your Secret Access Key.TOS_ENDPOINT: The endpoint for the TOS service (e.g.,https://tos-cn-beijing.volces.com).TOS_REGION: The region for the TOS service (e.g.,cn-beijing).
Optional for STS
TOS_SECURITY_TOKEN: If using a temporary token (STS), provide the session token here. The client will automatically use it if present.
Best Practices
- Error Handling: Always wrap SDK calls in
try...exceptblocks to handleTosClientErrorandTosServerError. - Parameter Validation: Validate parameters like
time,width, andheighton the client side before making an API call to prevent unnecessary errors. - Batch Operations: For
videoSnapshots, use a thread pool (likeThreadPoolExecutor) to perform multiple snapshot requests in parallel for better performance. Seescripts/video_snapshots.pyfor an example. - Credentials Management: Use a secure method to manage and refresh credentials, especially when using short-lived STS tokens.
Additional Resources
- For detailed parameters of each operation, see REFERENCE.md.
- For common end-to-end examples, see WORKFLOWS.md.
- For executable Python examples, see the
scripts/directory.
Installs
Security Audit
View Source
bytedance/agentkit-samples
More from this source
Power your AI Agents with
the best open-source models.
Drop-in OpenAI-compatible API. No data leaves Europe.
Explore Inference APIGLM
GLM 5
$1.00 / $3.20
per M tokens
Kimi
Kimi K2.5
$0.60 / $2.80
per M tokens
MiniMax
MiniMax M2.5
$0.30 / $1.20
per M tokens
Qwen
Qwen3.5 122B
$0.40 / $3.00
per M tokens
How to use this skill
Install byted-tos-video-process by running npx skills add bytedance/agentkit-samples --skill byted-tos-video-process in your project directory. Run the install command above in your project directory. The skill file will be downloaded from GitHub and placed in your project.
No configuration needed. Your AI agent (Claude Code, Cursor, Windsurf, etc.) automatically detects installed skills and uses them as context when generating code.
The skill enhances your agent's understanding of byted-tos-video-process, helping it follow established patterns, avoid common mistakes, and produce production-ready output.
What you get
Skills are plain-text instruction files — not executable code. They encode expert knowledge about frameworks, languages, or tools that your AI agent reads to improve its output. This means zero runtime overhead, no dependency conflicts, and full transparency: you can read and review every instruction before installing.
Compatibility
This skill works with any AI coding agent that supports the skills.sh format, including Claude Code (Anthropic), Cursor, Windsurf, Cline, Aider, and other tools that read project-level context files. Skills are framework-agnostic at the transport level — the content inside determines which language or framework it applies to.
Chat with 100+ AI Models in one App.
Use Claude, ChatGPT, Gemini alongside with EU-Hosted Models like Deepseek, GLM-5, Kimi K2.5 and many more.