The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from external OLX listings (titles, descriptions, parameters).
Ingestion points: Data retrieved from endpoints like /api/v1/olx/search and /api/v1/olx/offer as described in SKILL.md.
Boundary markers: The instructions explicitly direct the agent to "Quote those values as returned" and separate translations from the raw data to prevent instruction confusion.
Capability inventory: The skill is designed for data retrieval; it does not pipe external content into shell commands, file-write operations, or dynamic code execution.
Sanitization: The API provider performs automated server-side redaction, replacing phone numbers and email addresses with safety placeholders before the data reaches the agent.
[SAFE]: Sensitive authentication is managed correctly through the SCAVIO_API_KEY environment variable. The documentation provides placeholders for local setup rather than hardcoding actual credentials.
[SAFE]: External dependencies (scavio for Python/Node.js and @scavio/mcp-server) and network destinations (api.scavio.dev) are verified vendor resources belonging to the skill's author, 'scavio-ai'.

olx-api