The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The scripts scripts/fetch_article.py and scripts/generate_image.py contain code that explicitly disables SSL certificate verification using ssl.CERT_NONE. This bypasses standard security protocols, allowing an attacker to potentially intercept or alter data transmitted between the agent and external services (MitM attack).
[EXTERNAL_DOWNLOADS]: The scripts/fetch_article.py script is designed to download content from arbitrary URLs provided by the user. While intended for article analysis, this provides a mechanism for the agent to fetch and process potentially malicious remote content.
[DATA_EXFILTRATION]: The scripts/generate_image.py script communicates with the vendor's API at redfox.hk to generate images. This involves sending article fragments and prompts to a third-party service. Furthermore, the script attempts to read an API key from a sensitive file path in the user's home directory: ~/.qoder/apis/redfox.json.
[PROMPT_INJECTION]: The skill is vulnerable to Indirect Prompt Injection. It lacks boundary markers and sanitization when ingesting data from external URLs via scripts/fetch_article.py.
Ingestion points: External HTML content from user-specified URLs and user-supplied article text.
Boundary markers: Absent; there are no instructions to the model to ignore embedded commands in the source material.
Capability inventory: The skill can perform network requests and write files to the local system.
Sanitization: None; the scripts parse HTML and text structure without filtering for malicious prompt sequences.

visual-ops-writer