Engineering
Reviews, standups, incident response, architecture decisions, tech debt.
Engineering teams spend half their week on non-coding work: PR review, standup updates, on-call, architecture conversations, docs. Pace tightens each one. The engineering plugin works standalone (no required connectors), and pairs with the data plugin for the 'why is this slow' questions that show up in reviews.
Install for Engineering
The full engineering starter set is 3 plugins. Pick the install path that matches you.
Cowork (no terminal)
Click each link below; Cowork opens and asks you to confirm. Restart Cowork after the last one so connectors register.
Terminal (one command)
If you have claude on your PATH already, this installs the whole set in one go.
npx pace-tools install engineering data productivity pace-tools wraps claude plugin install; runs claude plugin marketplace add GoldenBerry-SO/Pace automatically if you haven't registered the marketplace yet.
Plugins to install
Pick from this set for the role. Primary plugins are essential; companions multiply value.
- engineering primary
Code review, architecture decisions, incident response, tech debt audits, deploy checklists, testing strategy, docs.
claude plugin install engineering@pace - data companion
When reviews touch performance, SQL, or analytics; pair with the data plugin to ground in real numbers.
claude plugin install data@pace - productivity companion
Standups, daily focus blocks, memory of past architecture decisions.
claude plugin install productivity@pace
Connectors to set up
Claude prompts to authorize each one the first time a relevant skill fires. You only do this once per project.
- (none required for engineering itself)
- Snowflake / Databricks / BigQuery (via the data plugin)
- Linear / Jira / Slack (via the productivity plugin)
Workflows
Common ways teams use these plugins day to day. Each one is a starting point; adapt the prompt to your context.
-
Pre-review code review
Runs a structural review before requesting human review. Catches the obvious stuff: dead code paths, missing tests, naming inconsistencies, broken contracts.
Review this PR. Look for real bugs first, then nits.
-
Standup summary
Generates your async standup by reading Linear tickets touched, git commits pushed, and what's blocking.
What should I say in standup? What is actually blocking me today?
-
Architecture decision
Pros/cons + recommendation on a system choice. Writes the ADR (architecture decision record) in your repo format.
Should we use NATS or Kafka for the new event bus? Walk me through the trade-offs.
-
Incident response
Triage playbook: diagnostic questions, likely causes, mitigation order, comms template. Useful at 3am.
P1: ingestion is lagging 40 minutes. Run an incident response.
-
Tech debt audit
Scans a directory or service and ranks debt by effort vs payoff. Surfaces stuff you keep meaning to fix.
Audit this directory for tech debt. What is brittle, what is stale, what should go first.
-
Documentation gap fix
Reads a module and writes the missing docs: contracts, edge cases, gotchas.
Document the rate-limiter middleware. What it does, how to tune it, gotchas.
Just say it
You don't have to memorize slash commands. After installing, type natural sentences into Claude and the right skill will fire. Each row is a real example.
- You say
Review this PR before I tag the team
Claude triggerscode-review - You say
Catch anything I missed before I open the PR
Claude triggerscareful-review - You say
Diagnose this race condition in checkout
Claude triggersdiagnose - You say
Write tests for the new checkout flow using TDD
Claude triggerstdd - You say
Spec the feature so we can ship it as stacked PRs
Claude triggersspec - You say
Commit my changes and open a PR
Claude triggersship-pr
Tips & tricks
Field-tested patterns from teams that have already shipped a quarter or two on these plugins.
-
Self-review before requesting review
`code-review` before you tag a human. Catches naming and silly mistakes; the human gets to focus on intent.
-
Pin the deploy-checklist
Use `deploy-checklist` before any prod push. Pin it (via the Pace router pin.mjs) once your team standardizes.
-
Testing strategy upfront
`testing-strategy` when scoping a feature, not when writing tests. Saves rewrite cycles.
-
Pair with /data for perf reviews
When the PR touches anything that hits the DB, run `write-query` against staging to verify the query plan before merging.
-
Incident response > heroics
`incident-response` has saved hours at 3am. Beats inventing a playbook under pressure.
-
ADR convention sticks
`architecture` writes ADRs in a stable format. Adopt the format and don't fight it; consistency is the point of ADRs.