openclaw - 💡(How to fix) Fix Feedback: improve local skill sync and feedback workflow

Feedback

While turning a repeated Gemini browser workflow into a reusable skill, I ran into a few OpenClaw workflow gaps that would be worth smoothing out.

What worked

• Workspace skills are discoverable once copied under the OpenClaw workspace skills directory.
• The copied skill became eligible, model-visible, user-invocable, and command-visible.

Friction points

1. There does not appear to be a first-class command for installing or syncing a local skill directory into the active OpenClaw workspace.

   • A command like openclaw skills install --local /path/to/skill or openclaw skills sync /path/to/skill would make Codex/OpenClaw skill reuse much easier.

2. I could not find a dedicated openclaw feedback command.

   • A small command that opens an issue template, writes to a feedback queue, or prints the canonical feedback URL would help users provide quick, structured reports from the CLI.

3. Skill validation is split across ecosystems.

   • It would be helpful if OpenClaw had a local skill validation/check command that catches required SKILL.md fields, executable scripts, and visibility/eligibility problems before copying into the workspace.

Why this matters

For long-running browser/UI workflows, users often iterate inside Codex first and then want to reuse the result from OpenClaw or Discord. A smoother local-skill install/sync/validate path would make that handoff reliable and less manual.

Environment

• OpenClaw version: 2026.5.7
• Platform: macOS