About ResearchZosho
ResearchZosho is an MCP server published by Wyrdsekai in the Search category: the research librarian for the rest of us: reads, keeps and questions your sources.
Recent commit activity is the strongest signal that reported issues will get attention.
Install
Claude Code
claude mcp add researchzosho -- npx -y @wyrdsekai/researchzosho-mcpnpx
npx -y @wyrdsekai/researchzosho-mcpThis server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.
Ask AI
✦
Ask AI about ResearchZosho
Powered by Claude · Grounded in docs
I know everything about ResearchZosho. Ask me about installation, configuration, usage, or troubleshooting.
Security checks
6 checks have not been run yet.
Releases
v0.5.5v0.5.5 · 8 Oct 2026ResearchZosho 0.5.5 New When a page a run needs has disappeared, it now reads the archived copy instead. If a page is gone, does not answer, or shows a sign-in wall in place of its text, the run fetches the copy that the Wayback Machine or…
v0.5.4v0.5.4 · 7 Oct 2026ResearchZosho 0.5.4 New Reports fix their own flagged sentences. After the checks, the writer gets back each sentence that was marked, with the reason and the sources, and rewrites it. A rewrite goes in only if it passes the same check;…
v0.5.3ResearchZosho 0.5.3 · 5 Oct 2026ResearchZosho 0.5.3 New YouTube research. A research run can search YouTube for channels and videos, check whether a channel still posts, read what a video says with the time of each line, and cite the moment a claim comes from. It needs…
v0.5.2ResearchZosho 0.5.2 · 1 Oct 2026ResearchZosho 0.5.2 Fixed Every MCP tool field with a fixed set of values (`sources`, `mode`, `op`, `confidence`, `reach` and the rest, 25 in all) now declares those values in its schema as an enum, where before they were named only in the…
v0.5.1ResearchZosho 0.5.1 · 29 Sep 2026ResearchZosho 0.5.1 This release fixes removing a report, which could delete claims you had added yourself, and filing under subjects, which asked the model about every claim in the library after each run. It adds `model use` to switch…