Add Executor local MCP skill example

[JGSphaela]

Summary

Reworks the Executor experiment away from DevSpace bridge tools and into a skill-based workflow.

This PR now adds an example Agent Skill at examples/skills/executor-local-mcp/SKILL.md that teaches the model to use the user's existing executor CLI through DevSpace's existing shell tools (bash or exec_command) to:

• inspect configured Executor sources
• search the Executor tool catalog
• describe unfamiliar tools before calling them
• invoke local MCP/API tools through executor call

It also documents how to copy the example skill into ~/.devspace/skills or another active skill directory.

Why this direction

The earlier bridge-tool approach worked as a proof of concept, but it duplicated what the local Executor CLI can already do from shell tools and added more model-facing tool names.

I also checked @executor-js/sdk; from its package docs/types it looks aimed at building or embedding Executor runtimes/plugins rather than acting as a thin client for a user's already-running local Executor service. So for this iteration, the skill approach keeps DevSpace core unchanged and avoids wrapping the CLI.

Validation

• npm run typecheck
• npm test
• npm run build
• Local Executor CLI smoke test:
  • executor tools sources
  • executor tools describe zotero.user.default.zotero_list_collections
  • executor call zotero user default zotero_list_collections '{}'

[coderabbitai]

Important

Review skipped

Draft detected.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 1c35e8ef-b1a7-4460-a29f-262b0fe2dafa

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[Waishnav]

tbh, that’s not how I would approach it. find out is there an executor SDK we can use for this implementation?

The current approach doesn’t really make sense to me because:

1. It adds extra tool calls and basically just wraps the local executor CLI.
2. The executor can already be called from the bash or exec_command tool, so we probably just need to educate the model through a custom skill for external MCP tools available through executor.

[JGSphaela]

That's fair, it just something I did for proof of concept and to get it working. I didn't know that there's a SDK already, I'll check it out!

[JGSphaela]

I scrapped the bridge-tool approach.

I checked @executor-js/sdk; from the package docs/types it looks more aimed at building or embedding Executor runtimes/plugins than acting as a thin client for a user's already-running local Executor service.

I pushed a new revision that keeps DevSpace core unchanged and adds an example skill instead. The skill teaches the model to use the existing executor CLI through bash/exec_command for local MCP/API integrations.