The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes local Python scripts (naver_search.py, naver_read.py, and naver_download_images.py) to perform web scraping and file management tasks.
[DATA_EXFILTRATION]: The skill initiates network requests to retrieve data from Naver. It implements a security control in scripts/_naver_http.py using the is_naver_url function, which validates that all requests are restricted to specific Naver domains (.naver.com, .naver.net, .pstatic.net), effectively preventing data exfiltration to unauthorized external domains.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external blog posts which may contain malicious instructions.
Ingestion points: naver_search.py (search snippets) and naver_read.py (blog body text) ingest external data into the agent's context.
Boundary markers: Absent; the skill does not wrap the external content in protective delimiters or provide explicit instructions for the agent to ignore embedded commands.
Capability inventory: The skill possesses the capability to write files to the local disk (naver_download_images.py) and perform network operations.
Sanitization: The scripts use regular expressions (TAG_RE) to strip HTML tags, but do not filter or validate the actual text content for potential prompt injection patterns.
[SAFE]: The image downloader script includes robust protection against path traversal. It uses os.path.realpath to verify that the target download path resides within the intended output directory, preventing the skill from writing files to sensitive system locations.

naver-blog-research