ds-implement
Phase 3 of /ds workflow. Execute analysis tasks with output-first verification.
npx skills add majiayu000/claude-skill-registry --skill ds-implement-edwinhu-workflows-3 --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
Orchestrates analysis tasks with mandatory visible output at every step. It enforces output-first verification, delegates each task to subagents via the ds-delegate template, and requires checking, logging, and verification before moving to the next task.
How it works
- States that every code step must produce visible output and be verified before continuing.
- Prohibits writing analysis code in the main chat; instead, uses a Task agent for each PLAN.md task according to the ds-delegate workflow.
- For each task in PLAN.md, dispatch a Task agent per the ds-delegate instructions, then read and verify the agent's output.
- If output is missing or incorrect, stop, log the issue, and re-verify after corrections.
- Log task completion to LEARNINGS.md and proceed to the next task until all PLAN.md tasks are completed.
- If triggered, load and apply ETL enforcement references when PLAN.md specifies ETL-related sections, and enforce SAS, data-quality, and boundary constraints as indicated.
- Continuously monitor context; if warning or critical, follow the ds-handoff or context-management protocols.
When to use it
Use when executing a multi-task DS implementation plan where visible outputs are mandatory after each step and orchestration is required to delegate to subagents for analysis tasks.
What it can touch
- Uses tools and hooks defined for the task orchestration, including Read and Agent interactions with the ds-delegate template. (See DS implementation workflow for delegation and verification steps.)
Caveats
- Instructions emphasize that main-chat code must not contain analysis logic; adherence to delegation protocol is mandatory. Violations trigger restart via Task agent.
- The flow relies on PLAN.md, LEARNINGS.md, and ds-delegate templates being available and correctly referenced.
## Overview Apply output-first verification at every step of analysis implementation. This is Phase 3 of the `/ds` workflow. ## Contents - [The Iron Law of DS Implementation](#the-iron-law-of-ds-implementation) - EVERY step MUST produce visible output - [Delegation](#delegation) - Main chat orchestrates, subagents analyze - [What Output-First Means](#what-output-first-means) - [Red Flags](#red-flags---stop-immediately) - [SAS Language Routing](#sas-language-routing) - Load SAS enforcement when PLAN.md specifies SAS - [Implementation Process](#implementation-process) - [Verification Patterns](#verification-patterns) - See `references/verification-patterns.md` - [Common Failures](#common-failures-to-avoid) - [Gate: Exit Implementation](#gate-exit-implementation) # Implementation (Output-First Verification) Implement analysis with mandatory visible output at every step. **NO TDD** - instead, every code step MUST produce and verify output. <EXTREMELY-IMPORTANT> ## The Iron Law of DS Implementation **EVERY CODE STEP MUST PRODUCE VISIBLE OUTPUT. This is not negotiable.** Before moving to the next step, you MUST: 1. Run the code 2. See the output (print, display, plot) 3. Verify output is
- Overview
- Contents
- The Iron Law of DS Implementation
- Delegation
- Delete & Restart Protocol
- What Output-First Means
- Rationalization Table
- Drive-Aligned Framing
- Red Flags - STOP Immediately
- Implementation Strategy Choice
- SAS Language Routing
- Implementation Process Flowchart
- Topic Change Protocol
- Implementation Process
What does the ds-implement skill do?
Phase 3 of /ds workflow. Execute analysis tasks with output-first verification.
How do I install it?
Run `npx skills add majiayu000/claude-skill-registry --skill ds-implement-edwinhu-workflows-3 --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.