vinaygiri

code-review

Full senior code-review capability for merge/pull requests, full-repository audits, single files/functions, and focused security or test reviews. Covers straight diffs, JS→TS ports, and full re-architectures (e.g. Cosmos→SQL). Enforces a two-stage find→validate discipline (every finding proven against real code, deterministic cases executed), a senior-reviewer dimension checklist, severity + evidence + type tags, and a fixed readable report format. Use whenever asked to review, audit, or critique code, or to produce a review comment on an MR/PR.

coordination

How the first-mate coordinator decomposes work, dispatches the crew across isolated worktrees (treehouse + tmux + gnhf via fleetmind-launch), monitors, consolidates, and delivers forge-agnostically (PR on GitHub / MR on GitLab via forge.sh). Use when orchestrating multi-task or multi-repo work across the agent crew.

doc-standards

Neutral, portable documentation templates and standards — design docs (HLD/LLD), ADRs, release notes, user guides, runbooks, API references, READMEs — plus rules for grounding every claim in code and adapting to a project's existing doc style/tooling (Markdown or Confluence, and its chosen diagram format). Use when writing or reviewing structured documentation.

engineering-standards

The senior/staff-level engineering baseline every Fleetmind agent operates by — rigor, verify-before-claim, no assumptions, minimal faithful changes, test-first, security & PHI awareness, honest reporting, and a clear definition of done. Preloaded by all crew agents; also useful whenever high-quality engineering discipline is needed.

java-spring-conventions

Senior Java + Spring Boot conventions, idioms, and pitfalls — dependency injection, transaction boundaries, layering, validation, exception handling, JPA/N+1, reactive (WebFlux), and testing. Use when writing, reviewing, or porting Java/Spring code. Loads on top of the project's own style — the project's conventions always win.

review-gate

Risk-tiered code-review gate for shipping changes. Use whenever a change is ready to push/merge, or when asked to "ship", "raise a PR", or "run the review gate". Enforces automated adversarial review plus human approval on medium/high-risk changes.

AI & Automation Listed

test-automation

Acceptance / UI / E2E test-automation conventions, Robot Framework first (keyword-driven .robot suites, Browser library) plus Playwright, and the discipline that keeps automated suites reliable — stable selectors, explicit waits (never sleeps), independent data-driven tests, tagging, and CI integration. Use when writing or reviewing automated acceptance/UI/E2E tests, especially Robot Framework robot scripts.

typescript-conventions

Senior TypeScript/Node.js conventions, idioms, and pitfalls — strict typing, safe null/error handling, boundary validation, async correctness, and testing. Use when writing, reviewing, or porting TypeScript or JavaScript (Node/Express/Nest/frontend). Loads on top of the project's own style — the project's conventions always win.

Web & Frontend Listed

ui-ux-design

Expert UI/UX design workflow and standards — token-first design systems, the shadcn/ui + Tailwind + Radix build stack, accessibility (WCAG AA), and self-contained HTML mockups. Use when designing interfaces, building front-end components, creating mockups/prototypes, or reviewing UI for quality and accessibility. The project's existing design system always wins.

Testing & QA Listed

e2e-testing

How to validate a change end-to-end (repro-first), not just with unit tests. Use when making or verifying code changes, fixing bugs, or when asked to prove a change works. Only load when actually changing/validating code.