The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to fetch and process content from arbitrary external URLs, creating a significant attack surface for indirect prompt injection.
Ingestion points: External web content is ingested via scrapling_fetch.py and the scrapling CLI.
Boundary markers: None. The skill does not implement delimiters or provide instructions to the agent to ignore potentially malicious commands embedded in the scraped data.
Capability inventory: The fetched content is returned directly to the agent's context, where it could influence subsequent logic or tool usage.
Sanitization: The scripts extract text or HTML but do not perform sanitization to strip potential injection patterns or malicious scripts.
[DATA_EXFILTRATION]: In scripts/scrapling_fetch.py, the fetch_http function uses verify=False when calling Fetcher.get. This explicitly disables SSL/TLS certificate verification, making the connection vulnerable to Man-in-the-Middle (MitM) attacks where an attacker could intercept or modify the scraped data.
[EXTERNAL_DOWNLOADS]: The scripts/scrapling_install.sh script performs a system-wide installation of the scrapling package and its dependencies using uv pip install --system "scrapling[all]". Additionally, it executes scrapling install, which downloads external browser binaries (Chromium) and system dependencies.
[COMMAND_EXECUTION]: The skill relies on executing shell scripts and Python wrappers that interface with the system environment to manage browser automation and network requests.

scrapling