The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes content from external websites via ironbee-browser-devtools-cli, which creates an indirect prompt injection surface where a malicious website could attempt to influence the agent's behavior.
Ingestion points: External data enters the context through take-aria-snapshot and take-ax-tree-snapshot commands in SKILL.md.
Boundary markers: The skill does not currently define explicit boundary markers or instructions for the agent to ignore commands found within the accessibility tree data.
Capability inventory: The skill is limited to the ironbee-browser-devtools-cli tool for browser interaction and reporting.
Sanitization: No explicit sanitization or filtering of the ingested accessibility tree content is described in the audit workflow.

accessibility-audit