xorio

Plugin xorio from marketplace xorio in radumarias/xorio-claude-plugin. Commands read from the repo's manifest — 25 days ago.

Skills (9)

• brainstorm-mythos

  Brainstorm-mythos [topic] [--target P] [--rounds N=2] [--lenses generic|optimization] [--out FILE] [+Nk]. All-Fable multi-agent fan-out (every agent on Fable with ultrathink max thinking): ideate across diverse lenses → adversarially refute → debate → synthesize a ranked report. Use ONLY when the us

• brainstorm

  Brainstorm [topic] [--target P] [--rounds N=2] [--mid-pct P=20] [--lenses generic|optimization] [--cross-model] [--fable] [--out FILE] [+Nk]. Multi-agent fan-out: ideate across diverse lenses (mixed STRONG/MID model tiers, mixed effort) → adversarially refute → debate → synthesize a ranked report. U

• cleanup-code

  Refactor and clean up code with structured review gates. Detects language, applies coding standards (Rust, TypeScript, Python), runs simplification and security audit. Primarily used as a sub-step of the polish workflow. Includes DRY analysis, Law of Demeter checks, and YAGNI violation detection.

• generate-tests-coverage

  Scan the entire project for under-tested code and generate tests to fill coverage gaps. Prioritizes critical business logic.

• generate-tests-module

  Generate unit and integration tests for a specific module, directory, or file. Analyzes existing code and creates idiomatic tests.

• polish

  Polish local code changes — simplify, clean up, review, and audit before PR. Runs code-simplifier, cleanup-code, code-reviewer, and security-auditor in sequence. Use when user says "polish my changes", "prepare for PR", "clean up before merging", or "make this PR-ready".

• review-loop

  Multi-round audit loop: /simplify → parallel finders (/code-review + optional /security-review + /review) → adversarial verify (3 lenses, default-refuted) → apply verified fixes → repeat until N rounds dry or MAX_ROUNDS. Use when the user wants a thorough quality + correctness pass on a branch/diff

• review

  Review local code changes with multi-agent PR review, code review, and simplification. Presents findings for user selection before fixing. Use when user says "review my changes", "check my code", "review before PR", or "code review my work".

• team-forming

  Team-forming <task-group file or inline spec> [--roles r1,r2,...] [--team-slug S] [--no-architect] [--no-review-plan] [--no-confirm-plan] [--no-explore] [--fable]. Compose and spawn an agent team for one task-group: an optional recon pass hands a distilled codebase map to a MAX-tier architect/planne

Slash commands (11)

• /xorio:check-deps

  Check if required external plugins and MCP servers are installed

• /xorio:commit-message

  Create a git commit

• /xorio:guide

  Interactive guide — discover xorio workflows, get recommendations, and start flows

• /xorio:mallware-check

  Scan folder recursively for malware, obfuscation, call-home, prompt injection, hidden/dormant loaders, and Claude Code plugin threats (read-only; for an untrusted clone run it from a parent directory, inside a VM)

• /xorio:polish

  Polish local changes — simplify, clean up, review, and audit before PR

• /xorio:preview

  Preview a Markdown or Mermaid file in a side panel; no path opens a file picker (needs the doc-preview mod)

• /xorio:review-pr-mythos

  Mythos all-Fable multi-agent PR review — every agent on Fable with max thinking; multi-lens review, deterministic + adversarial validation, verified fixes, looped to convergence. Use ONLY when the user explicitly says "mythos" or asks for the all-Fable / max-thinking variant; for any other deep PR r

• /xorio:review-pr

  Ultracode multi-agent PR review — multi-lens review, deterministic + adversarial validation, verified fixes, looped to convergence. This is the DEFAULT deep-review variant — prefer it unless the user explicitly says "mythos" or asks for the all-Fable variant (then use review-pr-mythos).

• /xorio:review

  Review local changes — multi-agent PR review, code review, and simplification

• /xorio:root-cause

  Identify the root cause of a problem using an evidence-grounded 5 Whys. `--deep` fans out parallel hypothesis branches with deterministic grounding and adversarial verification (multi-agent).

• /xorio:tests

  Generate tests — analyzes gaps, detects anti-patterns, generates idiomatic tests

Subagents (4)

• security-auditor

  Proactively scans for security vulnerabilities. Reviews code and infrastructure configurations against common security risks like the OWASP Top 10. <example>Context: User wants a security review. user: "Audit the auth module for security issues"</example> <example>Context: User is concerned about vu

• test-analyzer

  Internal agent for the /xorio:tests pipeline. Analyzes code to identify what needs testing, maps existing test patterns, and produces a prioritized gap list. Do NOT invoke directly — use /xorio:tests which provides the required standards content and scope. Output feeds into the test-generator agent.

• test-docs-advisor

  Internal agent for the /xorio:tests pipeline. Reviews project documentation and suggests improvements that would help with test generation quality. Do NOT invoke directly — use /xorio:tests which launches this agent in parallel with the test-analyzer when --no-docs is not set. - Glob - Grep - Read

• test-generator

  Internal agent for the /xorio:tests pipeline. Generates idiomatic unit and integration tests following project conventions and language-specific standards. Do NOT invoke directly — use /xorio:tests which provides the required gap analysis, standards content, and key file paths from the test-analyzer

Hooks (2)

• SessionStart

  cat ${CLAUDE_PLUGIN_ROOT}/rules/planning.md ${CLAUDE_PLUGIN_ROOT}/rules/tools.md

• PreToolUse

  node ${CLAUDE_PLUGIN_ROOT}/hooks/block-secret-file-reads.mjs

Similar plugins