The Agent Skills Directory

[SAFE]: The skill implements strong security controls for network operations. The scripts/wps_ocr.py script includes a URLValidator class that performs DNS resolution (socket.getaddrinfo) to verify that URLs provided by the user do not resolve to private, loopback, or internal IP addresses, effectively mitigating SSRF and DNS rebinding attacks. Additionally, it disables HTTP redirects (allow_redirects=False) during API calls to the trusted Kingsoft domain aiwrite.wps.cn.
[SAFE]: Credential management follows best practices by requiring the WPS_OCR_ACCESS_KEY to be provided via environment variables rather than hardcoding. The SKILL.md instructions correctly guide the user to store these in a local environment file.
[INDIRECT_PROMPT_INJECTION]: As an OCR tool, the skill is inherently susceptible to indirect prompt injection if the processed document contains malicious text instructions designed to influence the AI agent's subsequent behavior.
Ingestion points: User-provided file paths or URLs processed in scripts/wps_ocr.py via the --path or --url arguments.
Boundary markers: The script returns a JSON structure containing the extracted Markdown text but does not implement explicit boundary markers or 'ignore' instructions for the LLM.
Capability inventory: The skill has the capability to read local files, make network requests to the aiwrite.wps.cn domain, and return text content to the agent context.
Sanitization: No specific sanitization or filtering of the extracted text is performed before it is returned to the agent context.

wps-ocr