The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/citability_scorer.py utilizes the __import__ function to dynamically load the sys module for reading input from standard input (stdin) when a file path is not provided. This is a common pattern for command-line utilities but is categorized as dynamic execution due to the runtime module loading mechanism.
[EXTERNAL_DOWNLOADS]: The scripts/fetch_page.py utility enables the agent to fetch content from external URLs. It implements robust security features, including validation to block requests to non-global (private, loopback, or reserved) IP addresses to prevent Server-Side Request Forgery (SSRF) and ensuring that URLs do not contain embedded credentials.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function involves processing external web content, creating a potential attack surface for indirect prompt injection.
Ingestion points: Web pages fetched via scripts/fetch_page.py and local files scanned by scripts/brand_scanner.py.
Boundary markers: SKILL.md and references/skill-contract.md provide clear instructions to the agent to treat all scraped content as untrusted evidence and to disregard any instructions contained within that data.
Capability inventory: The skill includes network-read operations (fetch_page.py), local file-read operations (brand_scanner.py), and multiple file-write operations for report generation.
Sanitization: The skill relies on specific natural language constraints to maintain a trust boundary between agent instructions and ingested evidence.

geo-content-optimizer