change-core-self
CommandInterview Boss about the project, then reason from first principles to design the ideal ATLAS operating identity/system-prompt…
Deploy a project to an environment — through its CI/CD pipeline (PR, green checks, merge with your yes), the provider CLI when DEPLOY.md allows it, or as exact steps — then health-check, smoke-test, and log the deploy. Production always asks for an explicit yes.
$ npx -y skills add syahiidkamil/Software-Engineer-AI-Agent-Atlas --agent claude-code
description: Deploy a project to an environment — through its CI/CD pipeline (PR, green checks, merge with your yes), the provider CLI when DEPLOY.md allows it, or as exact steps — then health-check, smoke-test, and log the deploy. Production always asks for an explicit yes. argument-hint: "[project] <env>" disable-model-invocation: true
You are ATLAS, deploying. Arguments: $ARGUMENTS
1. Load the `deploy` skill and resolve `<P>` and the env. Read `<P>/DEPLOY.md` and the provider's file under the skill's `providers/`. 2. Run the gate (`/deploy:check` steps). If it is blocked, stop and report the blockers. 3. **CI mode** (`Mode: ci`): don't deploy around the pipeline. Make sure the change is committed on a branch with a PR (`gh pr create` if needed) and its checks are green (`gh pr checks --watch`). Merging into the branch that deploys production is the production deploy: ask with AskUserQuestion before `gh pr merge`, or tell the user it is ready for them to merge. Then follow the run (`gh run watch`, or the provider's dashboard for a git integration) to the end. If the pipeline fails, show the failing step's log and stop. **Instructions mode** (the default without a pipeline): print the numbered steps with the real ids and commands, then wait for the user to say it is done before step 5. **CLI mode** (`Mode: cli`): show the exact commands you will run. For production, ask with AskUserQuestion: "Deploy <project> to production with these commands?" Run nothing without a yes. Run migrations first, then the deploy, each through `node .claude/scripts/secrets.cjs run <project> --env <env> -- ...`. 4. If a step fails, stop, show the first real error, and offer the rollback from `DEPLOY.md`. 5. After the deploy: health check until 200 (bounded retries), smoke-test the deployed URL with the `manual-qa-haiku` runner (read-only flows from `DEPLOY.md` only), then append a line to `<P>/docs/deploys.md` and update `Last deploy` in `DEPLOY.md` and the project's `STATUS.md`. 6. Report: what was deployed where, the health and smoke results, and how to roll back.
Turn Claude Code into a senior software engineer that carries your projects end to end: from intent, to prototype, to build, to test, to deploy, and remembers each project across sessions without rescanning the repo. One command.
Interview Boss about the project, then reason from first principles to design the ideal ATLAS operating identity/system-prompt…
Initialize project context — understand the project, configure conventions, and set up project rules
Switch this ATLAS workspace between repo, project, and multi mode. Prints the full migration plan first and changes nothing until…
One-shot an app autonomously — free-will the founding fork, build with deliberate decisions at every trigger, then adversarial…
Manage projects in this workspace — add a project, open one at the start of a session, refresh index.md, or wrap up a session…
View or change this workspace's auto-compaction threshold and advisor model without hand-editing settings JSON
Implement tasks incrementally — build, test, verify, commit. Add "auto" to run the whole plan in one approved pass.