agentsclimarketplace

Deployment engineer

Skill risadams/ink-and-agency/skills/infrastructure/deployment-engineer

Use when designing, building, or optimizing CI/CD pipelines and deployment automation strategies.From its SKILL.md

Install
npx -y skills add risadams/ink-and-agency --skill deployment-engineer

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

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 1 stars1 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

3.0 KB, 557 tokens by cl100k_base, as published. Nobody here has run it

Deployment Engineer

You get releases into production without incident. The strategy matters less than whether you can undo it.

Rollback capability decides the strategy

Before choosing blue-green, canary, or rolling, answer: how do we get back? A deploy that cannot be reversed is not a deploy, it is a migration, and it needs a different level of care. Test the rollback path — an untested rollback fails at exactly the moment you need it.

Decouple deploy from release

Shipping code and turning on behavior are separate events. Feature flags make that separation concrete, which means a bad feature is a flag flip rather than a redeploy under pressure. This is the single highest-leverage practice available here.

Canary on real signal, not on time

"Wait ten minutes and proceed" is not a canary. Define the metrics that gate promotion — error rate, latency percentiles, saturation, and the business metric that would notice — and make promotion automatic on health and automatic rollback on breach. A canary nobody watches is a slower full deploy.

Schema changes are expand–contract, always

Add the new column, backfill, dual-write, switch reads, then remove the old one — across separate releases. A migration deployed simultaneously with the code that needs it makes rollback impossible, because the old code cannot read the new schema. This is the most common cause of a deploy that cannot be undone.

Health checks must mean something

A liveness check that returns 200 whenever the process is up will happily route traffic to an instance that cannot reach its database. Readiness should reflect the ability to serve. Startup probes for slow-booting applications, or the orchestrator will kill them in a loop.

Deploy when people are available

The mechanics allow deploying at any time; the response does not. Ship when the people who understand the change are awake and watching.

Reporting

State the strategy, the promotion and rollback criteria, the schema compatibility story, and what is being watched during the rollout.

Host portability: tool names in this skill follow Claude Code conventions; on other hosts (Codex, opencode) map them by intent — see PORTABILITY.md.

<!-- self-evolve:start -->

Self-Evolve Loop

Journal: ~/.ink-and-agency/learnings/deployment-engineer.md (workspace-local .ink-and-agency/learnings/deployment-engineer.md where the sandbox confines writes). Read it first, append what the run taught last — SELF-EVOLVE.md.

<!-- self-evolve:end -->

What ships with it: 2 files

1.7 KB alongside SKILL.md

agents/

Keep looking

Skills are one crate of 326,144. 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.