Add Claude Code plugin marketplace for ModelOpt agent skills

[kaix-nv]

What does this PR do?

Type of change: ?
New feature. Add Claude Code plugin marketplace manifest so ModelOpt agent skills can be discovered and installed with a single command.

Usage

# Add the ModelOpt marketplace (one time)
/plugin marketplace add NVIDIA/Model-Optimizer

# Install all agent skills (ptq, deployment, evaluation, orchestrator)
/plugin install modelopt-agent@NVIDIA/Model-Optimizer

This bundles all ModelOpt agent skills into one installable plugin:
- /modelopt-agent:ptq — quantize models with FP8, NVFP4, INT4
- /modelopt-agent:deployment — serve checkpoints via vLLM, SGLang, TRT-LLM
- /modelopt-agent:evaluation — benchmark accuracy via NeMo Evaluator Launcher
- /modelopt-agent:modelopt — end-to-end orchestrator (chains PTQ → deploy/eval)

Testing

Before your PR is "Ready for review"

Make sure you read and follow Contributor guidelines and your commits are signed (git commit -s -S).

Make sure you read and follow the Security Best Practices (e.g. avoiding hardcoded trust_remote_code=True, torch.load(..., weights_only=False), pickle, etc.).

• Is this change backward compatible?: ✅ / ❌ / N/A
• If you copied code from any other sources or added a new PIP dependency, did you follow guidance in CONTRIBUTING.md: ✅ / ❌ / N/A
• Did you write any new necessary tests?: ✅ / ❌ / N/A
• Did you update Changelog?: ✅ / ❌ / N/A

Additional Information

Summary by CodeRabbit

• New Features
  • Added a new Claude plugin package ("modelopt-agent" v0.1.0) providing model quantization (PTQ), multi-backend deployment support, evaluation tooling, and orchestration workflows.
  • Registered a marketplace entry ("modelopt-skills") so the plugin is discoverable and installable via the Claude plugin marketplace.

[copy-pr-bot]

Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually.

Contributors can view more details about this message here.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: c686af28-d858-4ca8-ae07-822179cf444f

📥 Commits

Reviewing files that changed from the base of the PR and between d7e59cc and 3a33af4.

📒 Files selected for processing (2)

• .claude-plugin/marketplace.json
• .claude/skills/.claude-plugin/plugin.json

✅ Files skipped from review due to trivial changes (2)

• .claude/skills/.claude-plugin/plugin.json
• .claude-plugin/marketplace.json

---

📝 Walkthrough

Walkthrough

Added two new Claude plugin configuration files: a marketplace manifest for modelopt-skills and plugin metadata for modelopt-agent. Both files include version 0.1.0, descriptive metadata about quantization, deployment backends, evaluation, and contact/repository details.

Changes

Cohort / File(s)	Summary
Claude plugin manifests ​.claude-plugin/marketplace.json, ​.claude/skills/.claude-plugin/plugin.json	Added a marketplace manifest registering modelopt-skills and a plugin metadata file for modelopt-agent (version 0.1.0) with descriptions (PTQ/quantization, deployment backends, evaluation), owner/author, homepage/repository, license, and keywords.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

🚥 Pre-merge checks | ✅ 4

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The PR title accurately summarizes the main change: adding a Claude Code plugin marketplace manifest for ModelOpt agent skills, which aligns with the file changes (marketplace.json and plugin.json) and PR objectives.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Security Anti-Patterns	✅ Passed	PR adds only JSON configuration files for Claude plugin marketplace integration with no Python code changes to modelopt package or examples.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch kaix/agent-marketplace

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[github-actions]

PR Preview Action v1.8.1
🚀 View preview at https://NVIDIA.github.io/Model-Optimizer/pr-preview/pr-1141/
Built to branch gh-pages at 2026-04-03 23:19 UTC. Preview will be ready when the GitHub Pages deployment is complete.

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In @.claude-plugin/marketplace.json:
- Around line 8-14: The plugin manifest declares "modelopt-agent" with source
"./.claude/skills" but the skills directory only contains plugin metadata and no
executable skill implementations, so commands like /modelopt-agent:* won't be
discovered; add actual skill files (e.g., TypeScript/JavaScript/Python handlers)
into the "./.claude/skills" folder that export the declared commands and
handlers referenced by the plugin manifest (ensure function/command names and
routes match the manifest and plugin.json), include proper entry points and
registration code so the runtime can discover commands, and update the package
manifest if needed to reference the new skill files.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 86e71c35-d8e4-4a64-b5fa-9358f048d2f0

📥 Commits

Reviewing files that changed from the base of the PR and between 48da281 and d7e59cc.

📒 Files selected for processing (2)

• .claude-plugin/marketplace.json
• .claude/skills/.claude-plugin/plugin.json

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 74.76%. Comparing base (df80a0f) to head (3a33af4).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1141      +/-   ##
==========================================
- Coverage   74.76%   74.76%   -0.01%     
==========================================
  Files         351      351              
  Lines       40072    40072              
==========================================
- Hits        29961    29958       -3     
- Misses      10111    10114       +3     

Flag	Coverage Δ
unit	54.74% <ø> (-0.01%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Edwardf0t1]

I think it would be better to add this after we’ve finished polishing the skills within our team.