agentleFS
Sign inStart free
lfrodrigueslfrodrigues/packt-claude-code-course-auth/.claude/skills/deploy/SKILL.md

deploy

Skill9 starsChanged 4 months agoMITView on GitHub

Deploy the current commit to production on Render, gated on a green verify and a prod smoke

  • Commits and pushes

155 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use pm-discovery, qa-phase and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • ww-w-aiSkill

    pm-discovery

    PM Agent Team - Automated product discovery, strategy, and PRD generation. Runs PM analysis workflow to produce comprehensive PRD before PDCA Plan phase.

    Also in 3 projects that use this

  • ww-w-aiSkill

    qa-phase

    Run QA tests at various levels (L1-L5) for features under PDCA management. Integrates with MCP tools to execute structured test plans.

    Also in 3 projects that use this

  • +10 more from ww-w-ai

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 139 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Then distrust the green: Read the API logs via the Render MCP and check for better-auth's "could not determine a client IP" warning.

    line 34
  2. Very specific:

    Do not re-enable autoDeploy to work around a failure here.

    line 48
  3. Very specific:

    Never report "deployed" from a queued or in_progress status.

    line 49

“Common” is measured: kept by at least 6 in 10 of the 139 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Deploy to production

What to deploy: $ARGUMENTS

Render's services are autoDeploy: false on purpose: a git push must not ship to production. This command is the deploy path, and the gate lives inside it.

  1. Refuse to deploy uncommitted work. git status must be clean, and the branch must be pushed — Render deploys what is on GitHub, not what is on your disk. If they differ, stop and say so.
  2. Run npm run verify. If it is red, STOP and report the failure. Do not deploy, do not "fix it quickly first", do not rerun hoping it passes. A red verify is the answer, not an obstacle.
  3. Find the service and trigger the deploy via the Render MCP (list_services → trigger_deploy).
  4. Watch until it is actually live (get_deploy / list_deploys). A deploy that reaches update_failed or build_failed is a failed deploy — report the status and the build log tail. Do not report success on a queued deploy.

  5. Smoke the deployed app, against the STATIC SITE's URL (that is the public origin; the API is reached through its /api/* rewrite):

    API=https://<static-site>.onrender.com npm run smoke

    6/6 or it did not work. Report the failing step verbatim.

  6. Then distrust the green. Read the API logs via the Render MCP and check for better-auth's "could not determine a client IP" warning. Smoke passes 6/6 whether or not per-client rate limiting works — a collapsed shared bucket still returns a 429, it just trips sooner. The suite cannot see the difference; the log line can. If the warning is present, say so plainly: the deploy is up and a security property is broken. See docs/plans/012-prod-deploy-and-mcp.md.

Policy:

  • There is no force flag and no skip-checks flag, deliberately. Break-glass is human-only: Render's dashboard → Manual Deploy. If you find yourself wanting to bypass step 2, that is the gate doing its job — tell the user what is red and let them decide.
  • Do not re-enable autoDeploy to work around a failure here.
  • Never report "deployed" from a queued or in_progress status. Wait, or say it is still running.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free