The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on executing bash commands for its primary functionality. It uses a tool named $B for browser interactions (listing, showing, and running scraping scripts) and a local binary at ~/.claude/skills/gstack/bin/gstack-learnings-log to record architectural insights and patterns discovered during the session.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from external websites.
Ingestion points: Web content is retrieved through $B html, $B snapshot, and $B text commands within the 'Prototype phase' described in SKILL.md.
Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from following instructions that might be embedded in the scraped HTML or text.
Capability inventory: The skill utilizes the Bash tool to perform network operations (via $B) and to execute local logging scripts.
Sanitization: The instructions do not specify any sanitization or validation of the retrieved web content before it is parsed by the agent to generate a JSON response.

scrape