The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes the official tencentcloud-sdk-python and cos-python-sdk-v5 libraries to communicate with Tencent Cloud services. These are well-known, trusted packages from a reputable service provider.
[INDIRECT_PROMPT_INJECTION]: The skill exposes a surface for indirect prompt injection through its media analysis features (e.g., in scripts/mps_av_understand.py). If an AI agent processes attacker-controlled media URLs, any instructions hidden in the media (such as OCR text or audio transcripts) could attempt to influence the agent's logic. This is an inherent risk of AI media analysis tools and is documented here for awareness.
[DYNAMIC_EXECUTION]: The scripts/mps_usage.py script employs importlib.util to dynamically load the local load_env.py helper module. This loading is performed from a static, local file path within the skill's directory and is used for environment configuration before API calls.

tencent-mps