rerun-discovery

Clean previous outputs and re-run Vern-Bot discovery pipelines via CLI.

15|2|Updated Feb 6, 2026
One-click install
npx skills add https://github.com/jdonohoo/vern-bot --skill rerun-discovery
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: rerun-discovery
Source: https://github.com/jdonohoo/vern-bot/tree/main/skills/rerun-discovery
Command: npx skills add https://github.com/jdonohoo/vern-bot --skill rerun-discovery

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Re-run an existing Vern-Bot discovery pipeline while removing prior generated output and preserving the original input so users can refresh analysis after prompt edits or configuration changes without manual cleanup.

Core Features & Use Cases

  • Project discovery and selection: Scans a chosen folder for projects containing input/prompt.md and helps the user pick the project to rerun.
  • Prompt review and edit flow: Shows the project's prompt, allows external edits, and re-reads the prompt before execution.
  • Safe CLI orchestration: Verifies the plugin root, builds a single vern-discovery CLI invocation with llm, pipeline, VernHole, and Oracle flags, performs targeted cleanup of output artifacts, and enforces long timeouts while avoiding execution from user-specified paths.
  • Use case: A developer updates the discovery prompt to pivot project direction and uses this skill to clean previous results and re-run the complete pipeline reliably.

Quick Start

Use the rerun-discovery skill to clean previous output and run a fresh discovery pipeline on './discovery/my-project'.

Frequently Asked Questions about rerun-discovery

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

FAQPage Schema
How do I rerun a discovery pipeline and clear previous output safely?

You can rerun a Vern-Bot discovery pipeline by using a skill that scans your folder for projects containing input/prompt.md, removes previous output artifacts, and re-executes the vern-discovery CLI with assembled flags.

What is the safest way to clean up discovery pipeline outputs before re-execution?

Safe cleanup targets generated output artifacts for removal while preserving the original input/prompt.md, verifying the plugin root, and enforcing long CLI timeouts to prevent partial pipeline execution.

Can I edit the prompt before re-running the Vern-Bot discovery workflow?

Yes, the workflow displays the project's prompt.md, allows external edits, and re-reads the prompt immediately before executing the vern-discovery pipeline to ensure the latest changes apply.

Does the vern-discovery CLI support LLM mode and VernHole flags?

Yes, the vern-discovery CLI supports LLM mode, single LLM, pipeline mode, VernHole, and Oracle flags, which the skill assembles into a single CLI invocation during pipeline execution.

Why should I not run vern-discovery directly from a user-specified path?

Running vern-discovery from user-specified paths is avoided to enforce safe execution practices, ensuring the plugin root is verified and preventing unsafe CLI orchestration of pipeline internals.