Principle-first software-development workflow for Claude Code — brief → plan → task → implement → review, built for current-gen models.
Agent
Reviews code with skeptical bug-hunting mindset: off-by-one errors, null references, async correctness, race conditions, query correctness, type safety. Used by /kenspc-task-review parallel review (Angle 4); also safe to invoke standalone with a project context.
Agent
INTERNAL: Part of /kenspc-task-review orchestration. Requires REVIEWREPORTS structured CONTEXT input from the calling skill — standalone invocation will fail the prerequisite check. Do not auto-delegate.
Agent
Reviews code for edge cases and error handling: null/empty/boundary values, error propagation, resource cleanup, server-side validation. Used by /kenspc-task-review parallel review (Angle 2); also safe to invoke standalone with a project context.
Agent
INTERNAL: Part of /kenspc-guide generation orchestration. Requires GUIDEPATH from a freshly generated guide document — standalone invocation will fail the prerequisite check. Do not auto-delegate.
Agent
INTERNAL: Part of /kenspc-plan generation orchestration. Requires PLANPATH from a freshly generated plan document — standalone invocation will fail the prerequisite check. Do not auto-delegate.
Agent
Reviews code quality: project conventions, readability, maintainability, naming, and structural issues. Used by /kenspc-task-review parallel review (Angle 3); also safe to invoke standalone with a project context.
Agent
INTERNAL: Part of /kenspc-task-review orchestration. Requires REVIEWREPORTS and ACCOUNTABILITYLIST CONTEXT — standalone invocation will fail the prerequisite check. Do not auto-delegate.
Agent
Reviews code completeness against requirements: are all required changes implemented, are there partially implemented features, do interfaces match design. Used by /kenspc-task-review parallel review (Angle 1); also safe to invoke standalone with a project context.
Agent
INTERNAL: Part of /kenspc-task generation orchestration. Requires TASKDOCPATH and SOURCEPATH from a freshly generated task document — standalone invocation will fail the prerequisite check. Do not auto-delegate.
Agent
INTERNAL: Part of /kenspc-task-implement orchestration. Requires validated task document path — standalone invocation will fail the prerequisite check. Do not auto-delegate.
Agent
Reviews test coverage and test quality: happy/edge/error path coverage, test correctness, behavior-not-implementation testing. Used by /kenspc-task-review parallel review (Angle 5); also safe to invoke standalone with a project context.