agentsclimarketplace

Suxiaoqiang cli

Skill AiGuangInc/suxiaoqiang-cli/skills/suxiaoqiang-cli

Use suxiaoqiang-cli (sxq) to sync, edit, preview and release Superun vibe coding projects from the terminal. Use when the user asks to pull/push Superun project code, publish a preview build, deploy/release a Superun app, check release status, or mentions "sxq", "suxiaoqiang-cli", or a Superun sessionId. 当用户要求同步/推送 Superun 项目代码、发布预览、 上线 Superun 应用、查看发布状态,或提到 sxq / suxiaoqiang-cli 时使用。From its SKILL.md

Install
npx -y skills add AiGuangInc/suxiaoqiang-cli --skill suxiaoqiang-cli

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

5.6 KB, ~1.3k tokens by cl100k_base, as published. Nobody here has run it

suxiaoqiang-cli (sxq)

sxq syncs a Superun project (identified by a sessionId) with a local directory, and drives the publish/release pipeline. All commands run non-interactively when needed — always prefer the non-interactive forms below; interactive prompts hang in agent environments (the CLI fails fast on non-TTY with a hint, but don't rely on prompts).

Prerequisites

  • sxq login requires a browser and must be done by the user. If any command reports "Not logged in / 未登录" or "credential expired / 凭证无效", ask the user to run sxq login themselves — do not attempt it. Exception: if the user hands you a token, run sxq login --token <token> (it validates the token and keeps the previous credential on failure). Never ask the user to paste a token proactively.
  • A project directory is bound via .sxq/config.json (created by sxq link). Check for it before assuming a directory is linked.

Core workflow

sxq link <sessionId> -y     # bind current dir to a project (verifies ownership; needs login)
sxq pull                    # pull remote files (incremental, three-way merge)
# ... edit files locally ...
sxq push -m "<summary>"     # push local changes (pulls first; aborts on conflict)
sxq publish                 # debug publish (preview build); polls until done, prints preview URL
sxq deploy -y -m "<log>"    # release to production; polls until live, prints live URL
sxq deploy --status         # read-only: pending/published versions + live URL

Command details & flags

  • sxq push [-m <message>] — pushes added, modified, and deleted text files. Respects .gitignore plus built-in ignores (node_modules, dist, .git, binaries >5MB are skipped). If it aborts with conflict markers (<<<<<<< local), resolve the markers in the listed files, then push again.
  • sxq publish [--message-id <id>] — asynchronous; the CLI polls up to 10 min. Success prints a preview URL. If it fails with an error message, report it to the user verbatim.
  • sxq deployreleases to production and may incur cloud service fees. -y skips the confirmation and acknowledges the fee. Do NOT pass -y unless the user explicitly asked to deploy/release. With no pending version it republishes the latest released version (no progress polling in that mode — verify afterwards with sxq deploy --status). Optional: --region CN|INTL for cross-region release, -m to override the changelog.
  • sxq pull — safe to run anytime; local-only edits are preserved via three-way merge. Conflicted files are listed and contain git-style markers; resolve before pushing.
  • sxq config set|get|unset|list — keys: host (API base URL), lang (zh/en).
  • --debug on any command prints full request/response logs (tokens masked) — use it when diagnosing failures.

Database migrations (sxq db push)

Superun projects use Supabase. Schema changes MUST go through migration files executed by sxq db push — never by pushing SQL files with sxq push (the CLI blocks any change under supabase/migrations/ during a normal push).

Full flow:

  1. Write the DDL in a new file under supabase/migrations/. The name MUST be <digits>_<memo>.sql (everything before the first underscore must be digits) — files not matching are SKIPPED with a warning, same as the Supabase CLI, so a misnamed migration silently never runs. Use a yyyyMMddHHmmss timestamp as the digits — generate it with date +%Y%m%d%H%M%S (e.g. 20260709120000_create_users.sql) — so execution order stays deterministic.
  2. Run sxq db push. It will:
    • pull remote changes first (aborts if there are merge conflicts — resolve, then rerun);
    • diff local files against the remote baseline to find migrations that are new;
    • execute the new migrations one at a time, in ascending timestamp order;
    • stop at the first failure and print the server's error message. Migrations before the failed one are already applied; fix the failing file and rerun — only the remaining (still-new) migrations execute again;
    • after success, the server stores each migration file as a project attachment automatically, and the CLI runs a final pull so the local manifest matches.
  3. Never edit an already-executed migration file — it is part of the remote baseline; write a new migration instead.
  4. Migration SQL should be idempotent where possible (create table if not exists, drop ... if exists).

Rules of thumb

  1. Run sxq pull before editing if the project may have changed remotely (e.g. the user also edits on the Superun web UI).
  2. After pushing code changes the user wants to see: sxq publish for a preview; only sxq deploy when they ask to go live.
  3. deploy --status is read-only and always safe for checking state.
  4. Exit code 0 = success. Non-zero exit prints an actionable error message on stderr — read it before retrying; do not blindly retry deploy.
  5. Never commit or expose the contents of .sxq/ (it contains the sessionId and session metadata; anyone with the sessionId may be able to read project files).

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 0 of the 12 instructions most ship operate skills give in ~1.3k tokens

Counted across 779 of the 1,178 authors here whose files we hold, read 2026-08-07

  • Document a rollback plan before deploymentin 41 of 779, across 22 files
  • Update the changelogin 21 of 779, across 19 files
  • Run the test suitein 20 of 779
  • Create an annotated git tagin 20 of 779
  • Clean up feature flags after full rolloutin 18 of 779, across 10 files
  • Verify deployment health after launchin 18 of 779, across 10 files
  • Test both feature flag statesin 17 of 779, across 9 files
  • Verify the working tree is cleanin 17 of 779
  • Make database migrations backward-compatiblein 16 of 779, across 8 files
  • Set up error monitoring before launchin 15 of 779, across 7 files
  • Monitor metrics at each rollout stagein 14 of 779, across 5 files
  • Create a GitHub releasein 14 of 779

Said here and by no other author read

  • prefer non-interactive command forms
  • ask the user to run login themselves
  • check for project config before assuming linkage
  • pull before editing if remote changed
  • publish for a preview
  • write schema changes as migration files

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 326,852. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.