The Agent Skills Directory

[COMMAND_EXECUTION]: The skill requires the agent to execute several administrative and data ingestion scripts using the Node.js runtime.
Verified script invocations: node scripts/ingest_signals.mjs, node scripts/ingest_trends.mjs, node scripts/file_report.mjs, node scripts/execute_decisions.mjs, node scripts/publish_airapp.mjs, and node scripts/setup.mjs.
These scripts are used for syncing data with a Busabase Space and deploying the internal AirApp UI, which is consistent with the skill's primary purpose.
[EXTERNAL_DOWNLOADS]: The skill declares dependencies on standard public packages from the npm registry.
Dependencies include busabase-sdk, hono, and @hono/node-server.
The use of busabase-sdk aligns with the skill's stated vendor integration.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external websites, such as competitor pricing pages, changelogs, and G2 reviews.
Ingestion points: Data enters the system via scripts/ingest_signals.mjs and scripts/file_report.mjs after the agent performs web searches or browser automation.
Boundary markers: The instructions do not define explicit boundary markers or "ignore embedded instructions" directives for the agent during the summarization of ingested content.
Capability inventory: The skill has the ability to execute shell commands (node scripts) and perform write operations to a Busabase data store.
Sanitization: While the UI code in content/busa-radar-app/app/app.js performs HTML escaping (escapeHtml), the Natural Language processing instructions do not specify sanitization or validation for the content being summarized.

busa-radar