Skip to content

Meta-Apply Patch Lander

Lands staged skill-corpus patches only after a fresh cross-model review and your approval.

Stars
18,482
Licence
MIT
Files
1 file

What it does

This skill is a gatekeeper for changes that other ARIS skills propose to the skill files themselves. Producer skills like meta-optimize can only stage candidate patches in a pending folder. When you explicitly type /meta-apply, this skill re-checks each patch you name by asking a different AI model family (via Codex) to judge the actual diff, refuses anything rejected, backs up the target file, applies the change, stamps a provenance record, and logs it. Any patch the reviewer rejects cannot be forced through.

Great for

  • Developers maintaining the ARIS research-automation skill corpus who run meta-optimize and want to review its proposals
  • Maintainers who want an audit trail of which model authored and which model reviewed each self-modification

Skip it if

  • Ordinary users who just want to do research or writing; there is nothing useful to land without staged patches from meta-optimize
  • Anyone expecting the stamp to prove a change is correct: the file says the provenance stamp is a process receipt, not a correctness claim, and reviewer judgment can be wrong
  • Use on a repository you have not backed up or under version control, since it edits skill files directly

Try asking

  • “meta apply 1,3”
  • “/meta-apply all”
  • “land the staged patches”
  • “应用优化”
Step-by-step guide: how to use Meta-Apply Patch LanderInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the MIT licence and an attribution note.

    meta-apply.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip meta-apply.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip meta-apply.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Review the REPORT from /meta-optimize before running this, since your invocation is the human landing approval.
  • Name specific patch numbers rather than 'all' so you land only what you have read.
  • Keep the backup folder under .aris/meta/backups/ and test the changed skill on your next run, as the skill itself recommends.

What's inside

  • SKILL.md

Redistributed unmodified from wanshuiyin/Auto-claude-code-research-in-sleep under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.