prompt-generator_skill

This skill generates structured phase-based implementation prompts using an orchestrator pattern and saves them to docs/prompts for subagent delegation.
  • Python

9

GitHub Stars

1

Bundled Files

3 weeks ago

Catalog Refreshed

2 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstart where the catalogue uses aiagentskills.

npx veilstart add skill mhylle/claude-skills-collection --skill prompt-generator

  • SKILL.md5.9 KB

Overview

This skill generates structured implementation prompts for phase-based project execution using an orchestrator/subagent pattern. It creates ready-to-run orchestration prompts, outputs them in chat, and saves them to the project's docs/prompts folder. The prompts encode delegation patterns, validation workflows, error handling, and handoff steps for each project phase. Use it to standardize multi-step implementations and feed downstream implement-plan tooling.

How this skill works

When invoked, the skill gathers phase variables (number, name, paths, project root, optional general plan and ADR directory). It reads a template, substitutes placeholders, and produces a complete orchestration prompt that instructs the main orchestrator to spawn and coordinate subagents. The generated prompt is displayed in chat and saved to docs/prompts/ using the project's save script following the phase-<N>-<name>.md naming convention.

When to use it

  • You need a ready-made orchestration prompt for a specific project phase (e.g., Phase 2: Data Pipeline).
  • You want to produce consistent delegation instructions for subagents across phases.
  • You need prompts that integrate ADR checks, plan updates, and handoff documentation.
  • You are preparing prompts for automated discovery by implement-plan or implement-phase.
  • You want standardized error handling, validation, and progress tracking built into prompts.

Best practices

  • Provide exact paths and the project root to ensure correct save and discovery behavior.
  • Include the general plan and ADR directory when available so the prompt can reference architecture decisions.
  • Use descriptive phase names (kebab-case friendly) to keep filenames discoverable by glob patterns.
  • Review generated prompts before execution to confirm delegation patterns and concurrency settings.
  • Archive completed prompts into docs/prompts/completed/ to maintain a clear implementation history.

Example use cases

  • Generate Phase 1 prompt for repository bootstrapping, instructing subagents to create base files and CI configuration.
  • Create Phase 2 orchestration prompt for a data pipeline with parallel file creation and sequential integration steps.
  • Produce a Phase 3 prompt that mandates ADR reads, records new ADRs during implementation, and updates the plan.
  • Batch-generate prompts for all planned phases so implement-plan can discover and schedule them automatically.

FAQ

Files follow docs/prompts/phase-<N>-<name>.md where <name> is a kebab-case phase name.

How does the prompt handle architectural decision records (ADRs)?

The prompt instructs the orchestrator to read referenced ADRs before implementation, create new ADRs for decisions made during work, and update the plan with ADR references when deviations occur.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational