PluginBench
Plugin
Maintained
MIT

xros

xros

Executable research OS: compile investigations into verifiable specs, falsify claims, and run multi-agent verification with mechanical checks.

What is xros?

xros is a research operating system that transforms vague investigations into falsifiable claims with verifiable methodologies. Use it to design rigorous research workflows, sharpen questions into testable hypotheses, and execute multi-agent verification engines that refuse to claim results without mechanical proof.

  • Compile long-horizon investigations into verifiable methodology specifications
  • Transform un-checkable questions into falsifiable claims with best-available checks
  • Run Tier-A/B multi-agent verification engine gated on real oracles
  • Reason through Tier-C claims with premises, pre-mortems, and dated tripwires
  • Refuse unverified claims and enforce mechanical validation requirements

How to install xros

Claude Code install command.

/plugin install xros@xiaolai

Works with

Agents this plugin has been detected to support.

Claude
Codex

Related plugins

nlpm

Active

Score, check, fix, and test natural-language artifacts across Claude Code, Codex, and Antigravity with tier-aware quality rules.

Claude
Codex
124
View repository →

cc-suite

Active

Bridge Claude Code, Codex CLI, and Antigravity CLI with unified agents and bidirectional delegation.

Claude
42
View repository →

north-star

Maintained

System prompt that removes confirmation-seeking, scarcity thinking, and best-practice ceiling constraints from Claude.

Claude
32
View repository →

English language coach that auto-corrects prompts and tracks improvement for non-native speakers

Claude
28
View repository →

bureau

Active

Turn AI sessions into a maintained, inspectable knowledge base with append-only logbooks, human-gated review, and dependency tracking.

Claude
22
View repository →

Auto-updated reference skills for Claude, Anthropic API, MCP, and platform features.

Claude
15
View repository →