MCP servers and skills: search, filter by where they run, and compare.
Multi-model peer review and drafting for articles, as a Claude skill + Python CLI. Claude models review blind, then answer each other's findings; a deterministic synthesis gives a scored report with a
SkillMITFew starsNo key
Multi-model, multi-vendor PR risk review as a Claude skill + Python CLI. Audits outbound statements, influence plans and hostile content per jurisdiction (TW/US/EU/CN): checklist every model answers,
SkillLink onlyFew stars
Offline source search, traceable evidence bundles and human review coverage, with optional read-only MCP tools.
PythonMIT★ 1Few starsNo key
JEV Gateway: PII/PHI guardrails for coding agents. Local MCP server, 0.8B inspector, and agent skills for Codex and Claude Code.
PythonMIT★ 1Few starsNo key
A self-hostable workbench for source acquisition, evidence preservation, retrieval and reviewed intelligence, with local inference and optional governed APIs.
PythonApache-2.0★ 1Few starsNo key
Shared room for Claude, Codex and Antigravity on one Windows PC: each agent woken in its own desktop app, local MCP daemon (experimental)
PythonMIT★ 1Few starsNo key
PyPi package for WhiteBox XAI MCP server https://www.whiteboxxai.com
PythonMITFew starsNeeds a key
Open-source, evidence-provenance-first scientific literature MCP server: 35 tools for literature search, citation/retraction checks, full-text quote verification, methodology audits, PRISMA-style scre
SkillApache-2.0Few starsNo key
Who stands behind it
