Back to Discover

claude-plugin-marketplace

artifact

xiaolai

Central marketplace for Claude Code plugins by xiaolai

View on GitHub
63 starsISCSynced Aug 2, 2026

Install to Claude Code

/plugin marketplace add xiaolai/claude-plugin-marketplace

README

claude-plugin-marketplace

Central marketplace for Claude Code and OpenAI Codex CLI plugins by xiaolai.

Two manifest files live in this repo:

  • .claude-plugin/marketplace.json — Claude Code entries (read by claude plugin install)
  • .agents/plugins/marketplace.json — Codex CLI entries (read by codex plugin install)

Each plugin repo can ship both layouts in parallel: .claude-plugin/ for Claude Code, .codex-plugin/ + codex/ tree for Codex. The two ecosystems coexist in one source repo per plugin.

Installation

For Claude Code:

claude plugin marketplace add xiaolai/claude-plugin-marketplace

For Codex CLI:

codex plugin marketplace add xiaolai/claude-plugin-marketplace

Available Plugins (Claude Code)

PluginDescriptionVersion
cc-suiteCC Suite — one plugin to bridge and delegate across Claude Code, Codex CLI, and Antigravity: single-source AGENTS.md, shared skills, mirrored hooks and MCP servers, full Claude↔Codex bidirectional delegation (including Codex reading Claude's session history), a Claude→Grok Build delegation lane over ACP, a bounded read-only Claude→Qwen Code review lane, project-scoped advisor agents, and opt-in bridging to more coding agents (Grok Build, opencode, Qwen Code, Kimi CLI). Supersedes cc-bridge and codex-toolkit.1.5.0
tdd-guardianTDD Guardian — test-first workflow, coverage gates, mutation testing, test quality audits0.7.2
echo-sleuthEcho Sleuth — mine past conversations, manage memory lifecycle, extract knowledge0.4.0
loc-guardianLOC Guardian — enforce per-file pure LOC limits with automated optimization strategies0.1.5
grillGrill — deep codebase interrogation with 6 specialized agents, 5 review styles, and 8 add-on pressure tests1.3.0
reason-grillreason-grill — deep argument interrogation: map a claim, then attack it from five adversarial angles (logic, evidence, counter, assumptions, integrity); a six-check calibration gate refuses to manufacture objections; returns a calibrated verdict (HOLDS / HOLDS-IF-NARROWED / WOUNDED / DEFEATED)0.3.0
docs-guardianDocs Guardian — documentation quality and freshness enforcer with staleness detection, accuracy checking, and auto-generation0.1.8
nlpmNLPM — score, check, fix, and test NL artifacts across Claude Code, Codex CLI, and Antigravity; tier-aware scoring with per-tool overlays (universal nlpm:conventions floor + conventions-claude, conventions-codex, conventions-antigravity overlays); per-tool Hooks tables (the three event vocabularies aren't 1:1 mappable); standalone Python validator (bin/nlpm-check) remains tool-agnostic for pre-commit hooks and CI; manifest-vs-disk consistency check still the differentiator in the Claude ecosystem1.2.1
claude-english-buddyEnglish language coach — auto-corrects prompts, translates non-English, tracks improvement over time0.5.1
mermaid-previewMermaid Preview — auto-preview diagrams in the browser on Write/Edit, offline-safe, dark-mode aware; opens the tab once per file0.1.4
ui-tokenizeUI Tokenize — block hardcoded UI values; rewrite-first PreToolUse hook corrects literals to design-token references on the way to disk; configurable strict / advisory strictness; per-project surfaces (by file kind) and ignore (by path) exclusion honored by hooks and audit; /tokenize:review dispatches the token-reviewer agent for semantic mis-pick review0.4.1
ui-responsiveUI Responsive — advisory responsive-design coach; flags off-catalog breakpoints, bare 100vh, fixed widths without max-width via PostToolUse additionalContext0.1.0
north-starNorth Star — 260-token system prompt overriding three RLHF-inherited presumptions (independence, calibration, first-principles); ambient + slash command + subagent layered delivery0.1.1
anthropic-docsAuto-updated Anthropic docs — 8 skills covering claude-code, claude-agent-sdk, anthropic-api, anthropic-platform-features, claude-connectors, claude-cowork, mcp-spec, plus anthropic-pulse (news + research digest); GitHub Actions pipeline refreshes daily from upstream docs + HTML feeds; supersedes the archived single-skill predecessor claude-agent-sdk-skill-autoupdated1.0.0
mac-it-guy-proPersonal IT guy for non-technical macOS users — checkups, safe cleanup, undoable file organising, evidence-first diagnosis, backups with a restore drill, home network, private internet access, automation that leaves reusable tools behind, and a tutor grounded in your machine's own numbers; memory that retires stale beliefs by retesting them; 159 test assertions1.9.1
eou-foundryEOU Foundry — recursive governance for Executable Operating Units (EOUs): faceted classification, generating-EOU constraints, ECP-governed change, no-self-approval. 12 skills covering candidate generation, audit, specify, refactor, promote, foundry-wide audit, ECP authoring, init scaffolding, Stage 0 capture, and judgment audit.0.8.1
bureauBureau — turn AI sessions into a maintained, human-reviewed knowledge base: capture → compile → review → query, gated by trust tiers (proposed → verified → canonical), with a BUREAU.md instruction (imported by CLAUDE.md) that makes every session honor them; renders to an offline gazette (built by the bundled press). A deterministic recursion engine tracks claim-level dependencies (rests_on + author-anchored spans) so changing an upstream claim flags every downstream page for review; a live Engine view (bureau:serve) shows three signals as you edit — dependency freshness (needs-review/stale), artifact currency (a verified file that drifted out from under a claim), and a convergence trend — and git-backed versioning renders any past board, diffs two versions, and pins named snapshots. Enables a crew of specialized agents. Self-contained.0.18.0
xrosXROS — executable research operating system. Four commands: compile a long-horizon investigation into a verifiable methodology spec; sharpen an un-checkable question (orient a newcomer in a field's vocabulary → falsifiable claim → best-available check, with its ceiling); run the Tier-A/B multi-agent engine (diverse routes, adversarial refutation) gated on a real oracle's exit code; or reason through a no-oracle Tier-C claim with premises, pre-mortem, and dated tripwires. Founding rule: no verification, no claim. Ships a bundled dependency-free validator + conformance suite.0.3.0
cc-bridgeDEPRECATED — use cc-suite instead0.1.0

Available Plugins (Codex CLI)

Codex ports are added incrementally as each plugin is converted. Status table — entries are removed from "pending" and added here once their .codex-plugin/ artifacts are committed and smoke-tested.

Most ports keep their skills under codex/skills/. xros instead ships a single root skills/ tree shared by Codex, Google Antigravity, and xAI Grok (all three implement the same skills/<name>/SKILL.md contract), with its .codex-plugin/plugin.json pointing at ./skills/. Antigravity and Grok are installed directly from the plugin repo (agy plugin install <git-url> / grok plugin install xiaolai/xros --trust), not through this marketplace — though Grok also reads Claude Code marketplaces automatically, so grok plugin install xros --trust works once this marketplace is registered.

PluginDescriptionVersionStatus
grillGrill — deep codebase interrogation with 7 specialized analysis skills, 5 review styles, and 8 add-on pressure tests1.3.0Install smoke-tested with Codex CLI 0.144.3
reason-grillreason-grill — deep argument interrogation: map a claim, attack it from five adversarial angles (logic, evidence, counter, assumptions, integrity) via $reason-grill-roast; a calibration gate refuses to manufacture objections; returns a calibrated verdict0.3.0Install smoke-tested with Codex CLI 0.144.3
eou-foundryEOU Foundry — turn messy workflows into auditable Executable Operating Units with recursive governance (12 skills: candidate generation, audit, specify, refactor, promote, foundry-wide audit, ECP authoring, init scaffolding, Stage 0 capture, judgment audit). v0.7.0 ships Stage 0 (captured_workflow + per-app domain_values constitutional layer + Rule 96 consumption). v0.8.0 ships agentic judgment (judgment_authorized flag, value_invocations trace, F14–F17 taxonomy, judgment_maturity J0–J4 axis, $audit-judgment skill, Rule 97, counterfactual-swap audit).0.8.1Install smoke-tested with Codex CLI 0.144.3
nlpmNLPM — reference-knowledge skills (17): the 50 Rules, the 100-point scoring rubric, per-tool conventions (Claude Code / Codex CLI / Antigravity), anti-patterns, vocabulary discipline, and authoring guides. Knowledge skills, not commands — $nlpm-rules + $nlpm-scoring to score an artifact, $nlpm-conventions-codex for the Codex layout. The interactive /nlpm:* linting commands stay a Claude Code plugin (they orchestrate sub-agents); cross-tool deterministic checks use the standalone bin/nlpm-check.1.2.1Install smoke-tested with Codex CLI 0.144.3; command/agent orchestration stays Claude-only
xrosXROS — an eXecutable Research Operating System. Four skills ($xros-compile, $xros-sharpen, $xros-run, $xros-reason): turn an investigation into a verifiable methodology spec, then gate the verdict on a real oracle's exit code. Founding rule: no verification, no claim. Ships a dependency-free validator + conformance suite.0.3.0Install smoke-tested with Codex CLI 0.144.3

Pending ports: tdd-guardian, loc-guardian, docs-guardian, echo-sleuth, claude-english-buddy. The Claude↔Codex delegation lane is now handled by cc-suite (single plugin, bidirectional). codex-guardian (Codex-artifact auditor) remains a planned future addition for the Codex-only audit niche.

Installing Plugins

Global (all projects)

claude plugin install cc-suite@xiaolai --scope user
claude plugin install tdd-guardian@xiaolai --scope user
claude plugin install echo-sleuth@xiaolai --scope user
claude plugin install loc-guardian@xiaolai --scope user
claude plugin install grill@xiaolai --scope user
claude plugin install docs-guardian@xiaolai --scope user

Codex installs

One-time setup (shell):

codex plugin marketplace add xiaolai/claude-plugin-marketplace

Codex registers this as marketplace name xiaolai.

Per-plugin install (shell):

codex plugin list --marketplace xiaolai --available
codex plugin add grill@xiaolai

You can also install from inside a Codex session:

  1. Start a session: codex (in your project directory)
  2. Type /plugins — opens the plugin TUI
  3. Find grill in the xiaolai marketplace, install/enable it

Invoking a plugin (inside a Codex session):

Codex uses the skill prefix $, not slash commands. Type:

$grill-roast

Or just describe the task in natural language ("do a multi-angle audit of this codebase") and Codex's auto-match will load the skill from its description.

(More entries to come as Codex ports land.)

Project only (current project)

claude plugin install cc-suite@xiaolai --scope project
claude plugin install tdd-guardian@xiaolai --scope project
claude plugin install echo-sleuth@xiaolai --scope project
claude plugin install loc-guardian@xiaolai --scope project
claude plugin install grill@xiaolai --scope project
claude plugin install docs-guardian@xiaolai --scope project

Scope reference

ScopeFlagEffect
User (global)--scope userAvailable in all projects (default)
Project--scope projectShared with team via .claude/plugins.json
Local--scope localLocal only, not committed to git

Managing Plugins

claude plugin list                           # List installed plugins
claude plugin update grill@xiaolai           # Update to latest version
claude plugin disable grill@xiaolai          # Temporarily disable
claude plugin enable grill@xiaolai           # Re-enable
claude plugin uninstall grill@xiaolai        # Remove

Troubleshooting

plugin install says "Plugin not found in marketplace 'xiaolai'"

The marketplace is a local git clone, and claude plugin install does not auto-refresh it before resolving the plugin name. If a plugin was added to the marketplace after your local clone was last updated, install will fail with a misleading "not found" error.

Refresh the marketplace, then retry:

claude plugin marketplace update xiaolai
claude plugin install <plugin-name>@xiaolai --scope user

This is a Claude Code CLI limitation, not a marketplace configuration issue. The plugin is genuinely listed in marketplace.json; your local copy is just stale.

Rendered live from xiaolai/claude-plugin-marketplace's GitHub README — not stored, always reflects the source repo.

19 Plugins

NameDescriptionCategorySource
cc-suiteOne plugin to bridge and delegate across Claude Code, Codex CLI, and Antigravity — single-source AGENTS.md, shared skills, mirrored hooks and MCP servers, full Claude↔Codex bidirectional delegation (including Codex reading Claude's session history), a Claude→Grok Build delegation lane over ACP, a bounded read-only Claude→Qwen Code review lane (Safe Mode + Plan mode + sandbox, isolated target copies, verified tool boundary), project-scoped advisor agents, opt-in bridging to more coding agents (Grok Build, opencode, Qwen Code, Kimi CLI) via a tool-profile registry, never-stale model selection (the config records the 'latest' policy), and a structured diagnostics engine (enabled-tools-aware health checks with verified auto-repair). Supersedes cc-bridge and codex-toolkit.developer-toolsxiaolai/cc-suite
tdd-guardianEnforces strict TDD discipline — test-first workflow, coverage gates, mutation testing, behavior-driven test quality auditsdeveloper-toolsxiaolai/tdd-guardian-for-claude
echo-sleuthMine past Claude Code conversations for values, decisions, mistakes, patterns, and wisdom. Manage memory lifecycle: audit staleness, extract knowledge, prune waste.developer-toolsxiaolai/echo-sleuth-for-claude
loc-guardianEnforce per-file pure LOC limits with automated optimization strategies — counts, flags violations, and provides opus-powered extraction plansdeveloper-toolsxiaolai/loc-guardian-for-claude
grillDeep codebase interrogation — multi-angle architecture review with 6 specialized agents, 5 review styles (including Paranoid Mode), and 8 add-on pressure tests. v1.3.0 ships the Codex CLI plugin layout alongside the Claude Code layout.developer-toolsxiaolai/grill-for-claude
reason-grillDeep argument interrogation — the grill method applied to reasoning. One command maps a claim (thesis agent), then attacks it from five adversarial angles (logic, evidence, counter, assumptions, integrity). A six-check calibration gate refuses to manufacture objections, so a sound argument mostly returns [SOUND]; returns a calibrated verdict (HOLDS / HOLDS-IF-NARROWED / WOUNDED / DEFEATED). v0.3.0 ships the Codex CLI layout alongside the Claude Code layout.developer-toolsxiaolai/reason-grill-for-claude
docs-guardianDocumentation quality and freshness enforcer — detect staleness, check accuracy, measure coverage, and auto-generate docs across any language and doc frameworkdeveloper-toolsxiaolai/docs-guardian-for-claude
nlpmNatural-Language Programming Manager — score, check, fix, and test NL artifacts across Claude Code, Codex CLI, and Antigravity. Tier-aware scoring with per-tool overlays. v1.0.0 marks the multi-tool branded release: nlpm:conventions splits into universal floor + per-tool overlays (conventions-claude, conventions-codex, conventions-antigravity); the scoring rubric splits Hooks per tool (the three event vocabularies are not 1:1 mappable); new artifact rows for .codex-plugin/plugin.json, .agents/plugins/marketplace.json, agents/openai.yaml sidecars, gemini-extension.json, LSP, monitors. Standalone Python validator (bin/nlpm-check) remains tool-agnostic and Claude-Code-free for pre-commit hooks and CI.developer-toolsxiaolai/nlpm
claude-english-buddyEnglish language coach for non-native speakers — auto-corrects prompts, translates non-English, refines with :: prefix, tracks improvement over time. v0.5.0: cleaner correction format — `wrong → right (reason)` per line, no-op suppression, category tagsdeveloper-toolsxiaolai/claude-english-buddy-for-claude
mermaid-previewAuto-preview Mermaid diagrams in the browser on Write/Edit/MultiEdit/NotebookEdit — offline-safe (inlined bundle), dark-mode aware, auto-reload on source change, per-file preview with LRU retention.developer-toolsxiaolai/mermaid-preview-for-claude
ui-tokenizeBlock hardcoded UI values from LLM-written code — rewrite-first PreToolUse hook silently corrects literals to design-token references on the way to disk; uncertain matches deny with structured suggestions (strict mode, default) or surface as PostToolUse advisories (advisory mode); per-project surfaces allowlist narrows which file kinds the hook scans, plus an ignore glob list excludes paths (test fixtures, generated or vendored code) from hooks and audit alike; /tokenize:review dispatches the token-reviewer agent to flag semantic mis-picks (right token vs. just any token); tokenize__* MCP tools for autonomous recoverydeveloper-toolsxiaolai/ui-tokenize
ui-responsiveAdvisory responsive-design coach — flags off-catalog breakpoints, bare 100vh, fixed widths without max-width, and stylesheets with no @media; surfaces findings as PostToolUse additionalContext, never blocks tool callsdeveloper-toolsxiaolai/ui-responsive
north-starA 260-token system prompt that overrides three structural presumptions every RLHF-trained LLM inherits from training: that you want confirmation, that old scarcity still applies, that best practices are ceilings.developer-toolsxiaolai/north-star-system-prompt
cc-bridgeDEPRECATED — superseded by cc-suite (github.com/xiaolai/cc-suite). Bridge a project's AI config across Claude Code, Codex CLI, and Gemini CLI.developer-toolsxiaolai/cc-bridge-for-claude
anthropic-docsAuto-updated multi-skill reference for the whole Anthropic doc ecosystem — claude-code, claude-agent-sdk, anthropic-api, anthropic-platform-features, claude-connectors, claude-cowork, mcp-spec, plus anthropic-pulse (news + research digest). 8 skills, refreshed daily from upstream docs and HTML feeds. Supersedes the archived claude-agent-sdk-skill-autoupdated (single-skill predecessor).developer-toolsxiaolai/anthropic-docs
eou-foundryRecursive governance for Executable Operating Units (EOUs). Faceted classification, generating-EOU constraints, ECP-governed change, no-self-approval. v0.7.0: Stage 0 (reference-grounded captured_workflow + per-app domain_values constitutional layer + Rule 96). v0.8.0: agentic judgment (judgment_authorized flag, value_invocations trace, F14–F17 taxonomy, judgment_maturity J0–J4 axis, $audit-judgment skill, Rule 97, counterfactual-swap audit).governancexiaolai/eou-foundry
bureauTurn AI sessions into a maintained, human-reviewed knowledge base: capture sessions to append-only logbooks, compile into consistency-checked cabinet dossiers gated by trust tiers (proposed → verified → canonical), and render an offline gazette (built by the bundled press). A deterministic recursion engine tracks claim-level dependencies (rests_on + author-anchored spans) via an append-only decision log, per-span revisions, an early-cutoff gate, and an fsck byte-fixpoint — so changing an upstream claim flags every downstream page for review. A live Engine view (bureau:serve) paints three signals onto the board as you edit — dependency freshness (needs-review/stale), artifact currency (a verified file that drifted out from under a claim), and a convergence trend (is the canon settling or thrashing) — and git-backed versioning renders any past board (build --at), diffs two versions, and pins named snapshots. Enable a crew of specialized agents or author your own. Karpathy LLM-wiki pattern + session provenance + a human review gate. Self-contained — no separate install.developer-toolsxiaolai/bureau
xrosXROS — an eXecutable Research Operating System. Four commands: compile a long-horizon investigation into a verifiable methodology spec; sharpen an un-checkable question (orient a newcomer in a field's vocabulary → falsifiable claim → best-available check, with its ceiling); run the Tier-A/B multi-agent engine (diverse independent routes, adversarial refutation) gated on a real oracle's exit code; or reason through a no-oracle Tier-C claim with premises, pre-mortem, and dated tripwires. Founding rule: no verification, no claim. Ships a bundled dependency-free validator and conformance suite.developer-toolsxiaolai/xros
mac-it-guy-proPersonal IT guy for non-technical macOS users — 12 plain-language commands, 4 agents, observation-first onboarding, a machine profile that retires its own stale beliefs by retesting them, a locked and transactional state writer, a PreToolUse guard that blocks destructive shell commands, a toolbox that turns repeated chores into reusable micro-tools and offers automations you did not know to ask for, learning maps built from your own measurements, and answers in your chosen language while everything on disk stays English. Ships a Codex layout with the same hooks. macOS only. 129 assertions.productivityxiaolai/mac-it-guy-pro

0 Comments

Login required
Log in to post a comment or update on this repo.

No comments yet — be the first to share an update.