[yujiachen-y]

What this adds or changes

New team stack stacks/jiachen with two skills. Builds on #1 (the marketplace); review the second commit.

Skill	What it does
lifting-loops	Finds the loops where the user keeps acting as a manual step for agents (retrying, judging, approving, carrying context, starting the same task by hand) and proposes fixes in order: delete a stale rule, change a skill / AGENTS.md, add a check, a skill or a scheduled task. Reads local MiniMax Code, Claude Code and Codex sessions; the ledger and proposals live in a gitignored private/ folder. Python standard library only.
making-content	Rewrites AI-sounding Chinese drafts in a restrained voice, makes Xiaohongshu image posts, and turns material into 60–180 s vertical videos (real-page captures, Edge or cloned voice, captions, text-density lint).

Both are copied from my own repository and released here under MIT. For this copy:

• personal samples stay out (private/ and test-cases/ are gitignored and were not copied);
• examples quoting my published posts, titles and account numbers are replaced with neutral examples;
• descriptions are single-line so check.mjs accepts them.

Each copied SKILL.md ends with an <!-- Upstream: ... --> line, and the stack README has an Origin section pointing to yujiachen-y/dotfiles, where the skills are maintained; the marketplace entry sets homepage to the same repository.

Also: marketplace entry, a row in stacks/README.md, a line in both READMEs, a README and a draft WORKFLOW.md. The icon is the template's for now.

Check output

$ MINIMAX_CODE_DIR=../minimax-code CHECK_BASE=origin/main node scripts/check.mjs
validate-plugin: stacks/jiachen passed the public Plugin V1 checks.
validate-plugin: stacks/ronny passed the public Plugin V1 checks.
validate-plugin: templates/stack-template passed the public Plugin V1 checks.
check: 3 stack(s) passed: stacks/jiachen, stacks/ronny, templates/stack-template

MiniMax Code's marketplace code reads the jiachen entry: .minimax-plugin manifest, version 0.1.0, both skills parsed with their full descriptions, install copy staged.

Real usage record

• lifting-loops: three weekly batch runs across two machines (2026-10-25 to 10-27) so far; the ledger tracks 36 loops, and of 14 proposals 10 were applied, 1 rejected, 3 are open. Applied proposals land as edits to my agent instructions and skills.
• making-content: used for my Xiaohongshu image posts and short videos since July 2026; three before/after samples are kept locally to calibrate the voice rules.

If this skill were deleted, where would the agent go wrong?

• lifting-loops: repeated corrections stay in my head; the same retry or approval keeps coming back in every session instead of turning into a rule or a script.
• making-content: drafts keep the usual AI tells (decorative emoji, contrast constructions, poetic endings), and videos come out text-heavy with no lint to catch it.

Checklist

• No keys, tokens, internal hostnames or private links
• Third-party content is credited with its license (music and fonts are fetched at run time with their credits)
• I agree to release this under the repository's MIT License

---

^{Need help on this PR? Tag @codesmith-bot with what you need.}