klank

Creates a new self-contained subagent identity in .claude/agents/ and its .github/agents/ (Copilot) and .junie/agents/ (Junie) mirrors. Use when introducing a new expert identity to the klank agent system.

Checks klank's agent setup for consistency - mirror parity, frontmatter, cross-refs, stale paths, descriptions. Use before any commit under .claude/, .github/agents/, or docs/agents/.

Diagnoses a failing CI/CD run via failed-step logs, debug logging, re-runs, flaky triage, and local repro. Loads a platform reference (e.g. github-actions). Use when a pipeline run is failing.

Audits and hardens an existing CI/CD pipeline against supply-chain and injection attacks. Loads a platform reference (e.g. github-actions) with a full checklist. Use when securing pipelines.

Keeps CI/CD pipelines healthy via dependency and runner-image updates, pruning, and reusable-pipeline versioning. Loads a platform reference (e.g. github-actions). Use for routine upkeep.

Scaffolds a secure-by-default CI/CD pipeline - least-privilege, pinned deps, caching, matrices, reusable parts. Loads a platform reference (e.g. github-actions). Use when creating a pipeline.

Makes a product design inclusive at the design stage - WCAG 2.2 POUR, inclusive design across impairments, an a11y checklist. Use when designing for or auditing accessibility.

Creates a product design as a structured design spec - user flows, information architecture, low-fidelity first, every state covered. Use when designing a new feature, flow, or screen.

Elicits requirements by questioning the user in exhaustive, batched rounds, then outputs a structured spec. Use when turning a vague feature idea or improvement item into a buildable spec.

Critiques an existing design or UI with a heuristic evaluation - Nielsen heuristics, cognitive load, hierarchy, consistency, plus a11y. Use when reviewing or auditing a design for rated fixes.

Writes product copy and microcopy - buttons, errors, empty states, labels - using a clarity-concision-consistency and voice-and-tone framework. Use when writing or fixing in-product wording.

Cleans up code at a chosen intensity/latitude over a scope (diff/area/repo) - simplify, de-cruft. Loads a stack reference. Use when tidying, not reviewing or strictness upgrades (develop-maintain).

Configures a project's type checker and build for strictness, correct module resolution, and fast builds. Loads a stack reference (e.g. typescript). Use when creating or tuning toolchain config.

Upgrades toolchain versions and migrates code to stricter types, removing escape hatches. Loads a stack reference (e.g. typescript). Use when bumping versions or migrating to stricter checks.

Ships a library so types and entry points resolve for every consumer, validated before release. Loads a stack reference (e.g. typescript). Use when packaging for publish or debugging resolution.

Reviews source code or a diff - correctness, least-code, type-safety, security, test adequacy - as severity-rated findings with fixes. Loads a stack reference (e.g. typescript). Use for code review.

Writes idiomatic, type-safe code - precise modeling, narrowing, boundary validation. Loads a stack reference (e.g. typescript). Use when writing or modifying source, not tidying (develop-clean).

Captures web page or running-app screenshots with Playwright - full-page, element, viewport variants - for visual QA, verification, and bug evidence. Use when a UI state needs capturing.

Exploratory and manual testing to find what automation misses - session charters, bug-hunting heuristics, every state and edge case. Use when probing a build by hand for defects before sign-off.

Verifies a change meets its acceptance criteria and is safe to ship - regression, bug triage, ship/no-ship call. Use for QA sign-off on behavior, not source-code review (develop-review).

Writes an effective automated test suite - testing pyramid, behavior over implementation, deterministic tests. Loads a stack reference (e.g. typescript). Use when adding or improving tests.

Evaluates and maintains third-party dependencies across their lifecycle - adoption vetting, health and security audits, updates, pruning. Use when adding or auditing a dependency.

Adds a Claude Code hook as a TypeScript file under .claude/hooks/ and registers it in .claude/settings.json. Use when a mechanically checkable constraint is currently only a documented bullet.

Creates a new procedure-skill with SKILL.md and a catalogue entry. Use when adding any new auto-triggered procedure to the klank agent system.

Runs the full NX build pipeline including TypeScript type-check and lint. Use before any PR or to verify structural changes.

Scaffolds a new NX library with correct Vite + Vitest config and @klank/* path alias. Use when adding a new shared library to libs/.

Starts the Vite dev server or full Tauri desktop app. Use when starting development or verifying UI changes in the running app.

Runs Vitest tests across the workspace or for a specific lib. Use after any code change before committing.

Upgrades pnpm workspace and Cargo.toml dependencies safely - detects breaking changes and runs tests. Use when dependencies need updating or a security advisory requires action.

Updates README and human-readable docs to reflect recent code or config changes. Use after any structural change - new lib, Tauri command, path alias, or route - to keep docs current.