Skill Claude CodeCodex
A guided entry point for adding, changing, removing, or promoting team rules for coding agents. It records where a rule applies, makes sure agents load it, and connects it to checks that can detect violations.
Skill Claude CodeCodex
A guided workflow for turning a project's test sandbox from a placeholder into a real test environment. A sandbox is the environment used to start, stop, inspect, and sometimes reset test runs.
Skill Claude CodeCodex
A coding workflow that coordinates an AI lead agent and, when useful, several worker agents to build features, fix bugs, refactor code, or migrate systems.
Skill Claude CodeCodex
Git workflow guidance for making changes to a Git repository, including branches, commits, merges, rebases, conflict resolution, pushes, and worktrees.
Skill Claude CodeCodex
A guided process for connecting a new or existing software project to harness-control, a system for organizing project rules, decisions, and execution entry points.
Skill Claude CodeCodex
A requirements workflow that turns a vague idea or existing system into user stories, acceptance criteria, a product requirements document, and optionally a prototype. A product requirements document describes what a product should do.
Skill Claude CodeCodex
A diagnostic process for finding and fixing gaps in harness-control itself, including its rules, skills, documents, evaluations, and context loading.
Skill Claude CodeCodex
An interactive workflow for turning product requirements into a technical design, including an interface contract when the system exposes an interface. It sits between requirements work and implementation.
Skill Claude CodeCodex
A test-production workflow that turns requirements or interface definitions into end-to-end tests, API tests, contract checks, or executable test scripts. It can also run an existing regression set; regression means checking that older behavior still works after changes.
Agent Claude Code
README from harness-base/harness-control.
Agent Claude Code
An API test-case writer that creates tests for interfaces such as HTTP, gRPC, or message events using an actual contract or other specified source.
Agent Claude Code
A skeptical reviewer for API test cases, tests that call software interfaces directly rather than clicking through a user interface. It checks coverage, source alignment, and the quality of normal, boundary, and error cases.
Agent Claude Code
An independent code reviewer that examines changes for bugs, security problems, technical debt, missing tests, and mismatches with documented contracts.
Agent Claude Code
A code-writing worker for a larger development workflow. It implements a specifically assigned part of a project while respecting the agreed design, interface contract, and file boundaries.
Agent Claude Code
A read-only reviewer that checks whether documentation was updated when related code or project-control files changed.
Agent Claude Code
An agent that writes browser-level end-to-end test cases for a requirement. End-to-end tests follow a user through a complete interface flow, such as opening a page, filling a form, submitting it, and checking the result.
Agent Claude Code
An independent reviewer for end-to-end test cases, which test a user journey through a complete application flow. It checks whether the cases truly cover normal, boundary, and error situations and match their source requirements.
Agent Claude Code
An independent evaluation agent that scores a completed task against specified prompts and a review rubric, a set of grading rules.
Agent Claude Code
An agent that creates a numbered feature list and maps each feature to user stories and exact sections of a product requirements document.
Agent Claude Code
An independent reviewer for checking how a software project is brought into a shared harness-control workspace. It examines the project setup, instructions, decision records, and verification entries without changing the files.
Agent Claude Code
An independent reviewer for product requirements documents, user stories, feature lists, and prototypes. It checks whether these planning materials are complete, consistent, traceable to agreed needs, and distinct from code review.
Agent Claude Code
A worker that writes a PRD, or product requirements document, from approved user stories. User stories describe what users need; the PRD turns them into defined scope, pages, flows, states, and observable acceptance checks.
Agent Claude Code
A coding-agent worker that builds clickable HTML prototypes for interface-related product requirements. It connects the main user flow with navigation, forms, validation, and state changes, using mock data instead of a backend.
Agent Claude Code
A requirements-gathering role that turns spoken or submitted information into a structured product summary.