A personal idea for a human agent workflow bundle where human and agents collaborate from the discovery through the implementation
maurizio100/human-agent-workflow-bundle
Skill Claude CodeCodex
Design, extend, and maintain arc42 architecture documentation in Markdown. Supports three operations: designing new chapters from scratch collaboratively, extending or editing existing chapters with new sections or content, and targeted mid-story edits invoked from implement-story. Use this skill whenever the user…
maurizio100/human-agent-workflow-bundle
Skill Claude CodeCodex
Collaborative backlog refinement — reads the current project state, identifies gaps, and suggests new stories to discuss with the user. Use this skill whenever the user wants to think about what to build next, asks for story suggestions, wants to review the backlog, or says something like "what should we work on…
maurizio100/human-agent-workflow-bundle
Skill Claude CodeCodex
Create lightweight chore stories for non-behavioral changes — layout tweaks, refactorings, dependency updates, tooling, or documentation. Use this when the change doesn't warrant Gherkin scenarios but still needs traceability. Trigger when the user says "quick fix", "chore", "refactor", "tweak", or describes a change…
maurizio100/human-agent-workflow-bundle
Skill Claude CodeCodex
Analyze finished Domain Storytelling stories (typically Egon .dst exports) and produce the downstream domain artifacts — bounded context candidates, glossary entries, relationship hints, and seed Gherkin stories. Use this skill whenever the user shares an Egon .dst file, a domain story write-up, a numbered…
maurizio100/human-agent-workflow-bundle
Skill Claude CodeCodex
Finalize implementation conventions after domain-storytelling and arc42-authoring are complete. Reads the architecture documentation, discusses conventions with the user, and writes the full CLAUDE.md plus layer-specific pattern files. Run this once arc42 content exists.
maurizio100/human-agent-workflow-bundle
Skill Claude CodeCodex
Author and refine Gherkin user stories as .feature files with proper IDs, epic grouping, declarative steps, and acceptance criteria suitable for triggering an agent-driven TDD loop. Use this skill whenever the user wants to write a new user story, promote a seed scenario from domain analysis into a real story, refine…
maurizio100/human-agent-workflow-bundle
Skill Claude CodeCodex
Implement a planned story — reads the plan from docs/plans/STORY-NNN.md on the current branch, writes failing tests, implements, runs the full test suite, self-reviews, and marks the PR ready for review. Assumes plan-story has already run and the plan file exists. Use this when picking up implementation after a plan…
maurizio100/human-agent-workflow-bundle
Skill Claude CodeCodex
Initialize a new project. Asks for a name and description, scaffolds the required folder structure, and writes a minimal CLAUDE.md — just enough to start domain modelling. Run once at project start.
maurizio100/human-agent-workflow-bundle
Skill Claude CodeCodex
Closes out an approved story — flips status, moves feature file, squash-merges PR, cleans up branches.
maurizio100/human-agent-workflow-bundle
Skill Claude CodeCodex
Plan a user story — read all relevant docs, produce the Plan + Test Plan artifact, write it to docs/plans/STORY-NNN.md and the draft PR body, create the story branch. Ends at the Plan human gate; does not touch any production code. Use this when the user wants to plan a story before implementing it, or when plan-story…
maurizio100/human-agent-workflow-bundle
Skill Claude CodeCodex
The shared output contract for domain-discovery analysis. Defines the domain artifacts (glossary, bounded-context files, the consolidated context map), where they live under docs/domain/, their exact formats and templates, the invariants that must hold, and the PROJECT.md update. This skill is NOT invoked directly by…
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: