The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches community sentiment data, weekly reports, and payment terms from external API endpoints.
Evidence: Requests are made to https://api.vibewatch.io and https://vibewatch-rc.up.railway.app via fetch and a shared x402 API client.
[INDIRECT_PROMPT_INJECTION]: The skill processes community-sourced data (from Discord, Telegram, X, and forums) which could potentially contain malicious instructions.
Ingestion points: Data is retrieved from the /api/v1/public/stacks-index sub-paths in vibewatch-sentiment.ts.
Boundary markers: The skill does not implement specific boundary markers or warnings to delimit external content for the agent.
Capability inventory: The script's capabilities are limited to data retrieval and console output via printJson. No file writing or arbitrary command execution capabilities were detected.
Sanitization: The skill outputs the retrieved data as JSON without specific sanitization of the underlying community-sourced text.
[OBFUSCATION]: The skill decodes Base64-encoded metadata to process payment receipts.
Evidence: The decodePaymentReceipt function in vibewatch-sentiment.ts uses Buffer.from(..., "base64") to decode the payment-response header, which is standard for the x402 protocol.

vibewatch-sentiment