Describe the work.
Discover the capability.
Type what you need an agent to do. Find the skills, agents, and tools that can do it.
Curated — domain-anchored, freshest first
Only entries whose URN authority matches the host that published them.
Runner twin of the 'archives' skill — the same tools executed server-side (no local setup). Tools: archives_extract_all, archives_extract_member, archives_list.
Runner twin of the 'arrow' skill — the same tools executed server-side (no local setup). Tools: arrow_read, arrow_schema, arrow_to_csv, arrow_to_json.
Runner twin of the 'avro' skill — the same tools executed server-side (no local setup). Tools: avro_sample, avro_schema, avro_to_csv, avro_to_json, avro_to_jsonl.
Runner twin of the 'dbf' skill — the same tools executed server-side (no local setup). Tools: dbf_schema, dbf_to_csv, dbf_to_json, dbf_to_text.
Runner twin of the 'doc' skill — the same tools executed server-side (no local setup). Tools: doc_to_csv, doc_to_json, doc_to_text.
Runner twin of the 'eml' skill — the same tools executed server-side (no local setup). Tools: eml_extract_attachments, eml_read, eml_to_csv, eml_to_json.
Give your agent the whole index
ARD isn't built into most clients yet. Add the StealthStack MCP server and your agent searches and lists every tracked resource automatically — discovery happens inline, no manual lookups. It's a discovery server (search_resources, list_resources, get_resource) over the whole crawled index.
Read the connection docshttps://stealthstack.ai/mcp