The Agent Skills Directory

[SAFE]: No malicious code, obfuscation, or security bypass attempts were identified. The skill's components are consistent with its stated purpose of marketing page analysis.
[EXTERNAL_DOWNLOADS]: The script scripts/conversion_audit.py uses urllib.request.urlopen to fetch content from external URLs. This is an intended functionality for auditing live web pages provided by the user.
[COMMAND_EXECUTION]: The skill involves the execution of a Python script to parse and score conversion signals in HTML content. The script uses standard libraries and does not perform dangerous system-level operations.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it is designed to ingest and analyze untrusted external data (HTML content from files or URLs).
Ingestion points: scripts/conversion_audit.py reads data via the --file or --url parameters.
Boundary markers: Absent; the agent processes the resulting analysis or raw content without specific delimiters to ignore embedded instructions.
Capability inventory: File read and network read (URL fetching) capabilities are present in scripts/conversion_audit.py.
Sanitization: The script extracts text for scoring but does not sanitize extracted content for potential LLM-targeted instructions.

page-cro