CBO
01Agent Claude Code
A senior technology-planning agent for designing systems, writing code, and reviewing technical decisions. It covers Python and Node.js development, automation pipelines, and technical feasibility checks.
Agent Claude Code
A senior technology-planning agent for designing systems, writing code, and reviewing technical decisions. It covers Python and Node.js development, automation pipelines, and technical feasibility checks.
Agent Claude Code
A final review agent for checking whether content matches the active brand identity and meets quality standards. Brand identity means the agreed voice, values, and style of a workspace.
Agent Claude Code
A content-production agent that turns ideas into complete drafts while following the active workspace’s brand voice. It supports analytical, professional, creative, and business-oriented writing styles.
Agent Claude Code
An agent role for financial analysis, return-on-investment checks, cost and risk reviews, and budget validation. ROI means comparing the expected gain from an action with its cost.
Agent Claude Code
A finance data-gathering and calculation agent that collects figures, organizes them, and calculates key performance indicators (KPIs), such as targets versus current results.
Agent Claude Code
A finance-review agent that checks financial analyses against fixed approval rules. It acts as a final gate before a financial decision is approved.
Agent Claude Code
A strategy-planning agent for finding bottlenecks, competitive advantages, risks, and high-leverage actions. A bottleneck is the constraint that most limits progress; leverage means getting a larger result from a smaller effort.
Agent Claude Code
A strategy-analysis agent that investigates the causes behind a situation rather than only describing its visible symptoms. It looks for bottlenecks, hidden variables, and actions where a small effort could produce a larger result.
Agent Claude Code
An execution planner that turns business analysis into a roadmap with milestones, actions, priorities, and measurable success targets.
Agent Claude Code
A technical architect and coding agent for designing systems, writing Python or Node.js code, building pipelines, and reviewing architecture. It follows rules for small dependencies, short functions, and handling data loss, network failure, and process crashes.
Agent Claude Code
A technical architecture agent that turns requirements into a system design and implementation specification. It decides module boundaries, data flow, dependencies, feasibility, and risks before coding starts.
Agent Claude Code
A coding agent that turns a technical specification into working code and reusable scripts. It also runs the code to check the result and records changes and test evidence in Korean.
Agent Claude Code
A final code review agent that checks whether code meets quality, security, reliability, and project-documentation rules. It gives only approval or a rejection with reasons.
Agent Claude Code
A marketing research and content agent that looks for current trends, competitors, customer reactions, and evidence in a chosen market. It turns the findings into a tested working idea.
Agent Claude Code
A marketing execution agent that turns market research into a concrete first draft, such as copy, an advertising script, a proposal, or code. It defines a success measure before execution and compares results with the original idea afterward.
Agent Claude Code
A market research agent that gathers current trends, competitor information, and examples of strategies or content that are working. Market intelligence means evidence about what is happening in a market now.
Command Claude Code
A command that records a user's explicit acceptance of a reviewed task. Acceptance means the user has taken responsibility for the result; passing tests or an agent's self-review does not count.
Command Claude Code
A command that moves an accepted task contract from the active folder to an archive folder without deleting it. A task contract is the record of the task and its agreed details.
Command Claude Code
A command for moving a completed handoff file into an archive folder. A handoff file records work passed from one person or agent to another.
Command Claude Code
A read-only diagnostic command for checking whether HamesSystem rules, files, paths, hooks, and environment mirrors are complete and consistent.
Command Claude Code
A command that starts a task marked READY after the current user explicitly requests it. It checks the task's contract, records the approval, and changes the task to ACTIVE.
Command Claude Code
Create a model handoff using a single canonical handoff document and then validate it.
Command Claude Code
A content and index audit command for checking workspace documents and their links. An index is a catalogue that records what files belong to a workspace.
Command Claude Code
A command that limits a session to a chosen workspace, meaning a project directory or group of directories. It can also load that workspace’s CLAUDE.md file, which contains local working rules.