sdlc-implement-plan

Coordinates isolated worker agents to implement locked feature plans with ordered integration and validation.

7|5|Updated Jul 22, 2025
One-click install
npx skills add https://github.com/nebius/nebius-ps-services --skill sdlc-implement-plan-nebius
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: sdlc-implement-plan
Source: https://github.com/nebius/nebius-ps-services/tree/main/skills/sdlc-implement-plan
Command: npx skills add https://github.com/nebius/nebius-ps-services --skill sdlc-implement-plan-nebius

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and assets (resource) components.

What problem does it solve? Implementing a locked feature plan with multiple parallel tasks risks shared branches, conflicting worktrees, hidden failures, and scope creep. This Skill coordinates dependency waves of isolated worker agents so each task gets its own branch, worktree, and verified commit before ordered integration. ## Core Features & Use Cases - Isolated worker dispatch: Each task runs in a fresh ephemeral agent (Codex or Claude) with its own scope cwd, workspace-write sandbox, schema-bound output, and digest-protected assignment. - Ordered wave integration: The coordinator verifies each worker result, creates one direct-child commit per task, merges in stable order, runs combined validation, and performs non-force cleanup. - Corrective repair flow: Diagnosis-bound corrective assignments rerun the original regression oracle first, with digest-bound proof required before integration. - Use Case: After sdlc-tdd produces red tests for a locked plan, the coordinator dispatches wave tasks to isolated workers, collects typed spec_gaps instead of silent spec edits, and advances the run state to implemented only when every wave is validated and cleaned. ## Quick Start Ask the Agentic SDLC coordinator to implement the current locked feature plan using sdlc-implement-plan after the TDD phase is complete.

Frequently Asked Questions about sdlc-implement-plan

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I implement a locked feature plan with parallel AI agents?

Use this Skill within the Agentic SDLC workflow after the TDD phase. It dispatches one fresh isolated agent per task with its own branch and worktree, then the coordinator verifies results, commits, and merges waves in stable order.

How does the sequential worker fallback work without native agents?

When native isolated agents are unavailable, the private launcher runs one ephemeral codex exec or claude --print process per assignment with exact scope cwd, workspace-write sandbox, stdin instructions, and schema-bound JSON output.

Does sdlc-implement-plan work with both Codex and Claude Code?

Yes. It supports Codex via $sdlc-implement-plan and Claude Code via /sdlc-implement-plan or the plugin form, using each host's native invocation syntax and ephemeral session flags.

What happens when a worker discovers a requirement or design gap?

Workers never edit canonical specs. They return typed spec_gaps in the schema-bound result, which sets status to replan_required and routes reconciliation back to the root coordinator before any task commit.

When should I not use sdlc-implement-plan?

Do not use it to edit requirements, designs, or locked plans, broaden scope, remove tests to pass, or reopen sealed or completed execution. Plan or design defects route back to sdlc-create-plan or sdlc-create-design.