Pre-Deploy Skill
<scripts>Cross-stack pre-deploy validation and remediation for node/rust/python.
Available scripts
| Script | Purpose |
|---|---|
scripts/stack-detect.sh | Detect project stacks (node, rust, python) |
scripts/deploy-check.sh | Full verification (lint/type checks + build) |
scripts/lint-check.sh | Run stack-specific quality checks |
scripts/build-run.sh | Run stack-specific build phase |
scripts/pm-detect.sh | Node package manager detection (npm/yarn/pnpm/bun) |
Workflow
Full validation (recommended)
scripts/deploy-check.sh
Step-by-step validation
# 1) quality checks
scripts/lint-check.sh
# 2) build phase
scripts/build-run.sh
Stack behavior summary
- Node.js: typecheck/lint (if configured) + build script (if configured)
- Rust:
cargo fmt --check+cargo clippy+cargo check+cargo build --release - Python: lint (
ruff/flake8if available) + type/syntax check (mypyorcompileall) + build (poetry buildorpython -m build, fallbackcompileall)
| Category | Required |
|---|---|
| Thinking | Use Sequential Thinking (3-5 steps per issue) |
| Tracking | Track failures with TodoWrite |
| Strategy | Run stack checks -> fix sequentially -> rerun checks -> run build |
| Validation | Re-check affected targets after each edit |
| Completion | Confirm deploy-check passes before completion |
<execution_plan>
- Run
scripts/deploy-check.sh - Convert errors into TodoWrite items by stack and priority
- For each issue:
- Run sequential-thinking
- Apply focused fix
- Re-run relevant check
- Re-run full deploy check
- Report final status and remaining risks
</execution_plan>