By jjjona
Artifact-first Research → Plan → Implement workflow. Skills write durable docs to 00-docs/ so work resumes across fresh sessions.
Capture a bug report or feature request as a structured issue file under 00-docs/01-issues. Use when the user reports a problem or asks to "create an issue for ...".
Implement an approved plan from 00-docs/03-plans — execute phases in order, pausing for human verification after each phase (don't chain phases unless told to). Use when the user asks to implement, execute, or follow a plan, or says "start phase X", "continue with the plan", "continue with phase X".
Create a detailed, testable, phased implementation plan through research and iteration — working interactively, with explicit sign-off on the phase breakdown before writing the full plan (never dump a complete plan in one pass). Use when the user asks to plan, create a plan, or wants a structured plan before coding.
Document and explain an existing codebase without critique or recommendations. Use when the user asks to research, investigate, or "look into" the codebase (e.g. "research how auth works", "look into X").
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
ds Claude Code pluginAn artifact-first Research → Plan → Implement workflow for Claude Code, packaged as the ds plugin (skills invoked as /ds:research, /ds:plan, …).
00-docs/), so work is resumable without chat history.Good research + a good plan = good code. A bad plan = exponential pain during implementation.
/plugin marketplace add https://github.com/jjjona/agentic-work
/plugin install ds@ds-tools
| Skill | Use it when |
|---|---|
/ds:research | "Research how authentication works" — documents the current codebase, no changes. |
/ds:plan | "Create a plan to do X, using 00-docs/02-research/..." — interactive, phased, testable plan. |
/ds:implement | "Implement phase 1 of 00-docs/03-plans/..." — executes a phase and verifies it. |
/ds:create-issue | "Create an issue for this bug ..." — captures a structured issue brief. |
Skills trigger implicitly from how you phrase the request, or explicitly as /ds:research, /ds:plan, etc.
00-docs/02-research/00-docs/03-plans/Between major steps, start a fresh session and anchor on the artifact, e.g.
"Plan the changes from 00-docs/02-research/2026-01-24-1530-auth-flow.md."
Each step may take several passes, so iterate.
/ds:create-issue is optional; use it to capture a bug or feature first, then feed it into research.
00-docs/)Created in your target project at runtime (the plugin does not ship them):
00-docs/01-issues/ — issue briefs00-docs/02-research/ — current-state writeups00-docs/03-plans/ — phased, testable plans00-docs/04-progress/ — short memos to resume work laternpx claudepluginhub jjjona/agentic-work --plugin dsHarness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
Comprehensive C4 architecture documentation workflow with bottom-up code analysis, component synthesis, container mapping, and context diagram generation
Superpowers Plus core skills library for Claude Code: planning, execution routing, TDD, debugging, and collaboration workflows
Unity Development Toolkit - Expert agents for scripting/refactoring/optimization, script templates, and Agent Skills for Unity C# development
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
Persistent file-based planning for AI coding agents. Crash-proof markdown plans (task_plan.md, findings.md, progress.md) that survive context loss and /clear, with an opt-in completion gate and multi-agent shared state. Manus-style. Works with Claude Code, Codex CLI, Cursor, Kiro, OpenCode and 60+ agents via the SKILL.md standard. Includes Arabic, German, Spanish, and Chinese (Simplified and Traditional).