testdouble/han

Architectural Analysis examines a named module, directory, or feature and produces a structured architecture report. Use it when you need design, behavior, risk, and boundary guidance without changing code.

Automated Test Planning analyzes source code to identify coverage gaps, edge cases, and the tests a branch needs. Use it when you need a standalone test plan for files or directories, not manual test steps or test code.

Code Overview produces a progressive-disclosure summary of unfamiliar code or a pull request’s changes, explaining why it exists, what it does, and where to start. Use it to get oriented before working on or reviewing code; it writes only a scratch file and does not change the target.

Code Overview To Confluence turns unfamiliar code or a pull request into a progressive-disclosure overview and publishes it to a Confluence location you specify. Use it when you need code explained and posted for review or sharing.

Code Review guides an AI agent through a structured review process for changes. Use it to assess diffs for bugs, regressions, test gaps, YAGNI, and severity-ranked findings.

Code Walkthrough walks a person through code changes one step at a time, starting at the entry point and following the flow that changed. It is used to learn, teach, or prepare to review existing code before extending it.

Coding Standards creates and updates project-specific conventions, rules, and guidelines. Use it when you need to draft, convert, or revise standards for a codebase.

Create A New Skill walks you through building a Claude Code skill with /skill-builder, from framing the request to writing the skill files and review notes. Use it when you want to add a deterministic skill to an existing or new plugin.

Design An API defines the contract for an API or interface change inside one codebase. Use it when you need to shape props, signatures, parameters, payloads, or module boundaries without writing code.

Documentation Update updates Han plugin documentation so skills, agents, indexes, and cross-references stay current. Use it after changing Han docs or related guidance, and it scopes itself to the touched branch or runs a full sweep on the default branch.

Edit For Readability rewrites an existing draft or file so the main point comes first, paragraphs carry one idea, and sentences stay short and active while preserving every fact. Use it to polish prose for readability without changing code, diagrams, or citation identifiers.

Explanation Guidance helps you write escalations, confirmation turns, and stop prompts as outcomes a non-implementer can observe. Use it when explaining a technical consequence or asking for input from someone who will not open the code.

Han Release cuts a Han release by updating the changelog, tagging changed plugins, and publishing release notes. Use it when shipping a new Han version or preparing tags and a release draft.

Investigate traces software issues from symptoms back to root cause using evidence from code, logs, and tests. Use it when debugging bugs, failed integrations, or API behavior that needs a deep dive rather than a code review.

Investigate To Confluence runs an evidence-based investigation of a bug, failure, or unexpected behavior and publishes the resulting report to a chosen Confluence location. Use it when you want debugging findings posted to a Confluence page or space.

Manual Test Planning produces a plain-language manual test plan from provided context, including a summary, named tests, and step-by-step expected outcomes. Use it when you need acceptance walkthroughs, hand-tested QA steps, or a plan for a feature, change, branch, or pull request.

Markdown To Confluence publishes an existing local Markdown file to a specified Confluence page or space through the Atlassian MCP server. Use it when you want to create or update a Confluence page from Markdown without rewriting the document.

Pairing builds work collaboratively in small, reviewable pieces so the person stays in the lead and can steer each step. Use it for paired coding, design decisions, or writing when you want review before moving on.

Plan A Feature To Confluence builds a feature specification with plan-a-feature and publishes it to a chosen Confluence location. It posts the spec as a parent page and related planning artifacts as child pages.

Project Discovery scans a codebase to identify its languages, frameworks, layout, and run commands, then writes a short reference into AGENTS.md or CLAUDE.md. Use it when you need a concise map of where things live and how to work in the project.

Project Documentation To Confluence creates project documentation for a feature, system, or component and publishes it to a Confluence page or space. Use it when you need documentation drafted, reviewed, and sent to a specific Confluence destination.

Readability Guidance surfaces Han’s shared readability standard and writing voice into the caller’s context before drafting prose. Use it when a prose-producing skill needs the standard available for self-checking and voice alignment.

Refactor restructures existing code without changing behavior using a test-gated loop. Use it to clean up, simplify, or reshape code after tests are green and the target’s behavior is covered.

Test-Driven Development writes code through a test-first red-green-refactor loop with an observed-failure gate. Use it to implement features behavior by behavior from tests, including TPP or ZOMBIES ordering.

Work Items To Jira breaks an existing work-items.md file into Jira tickets in one project, creating one issue per slice and linking dependencies. Use it after planning work items when you want implementation tickets tracked in Jira.