reversa-challenger

Runs premortem analysis on ideation options and writes adversarial risks.md documents.

1.5k|405|Updated Apr 26, 2026
One-click install
npx skills add https://github.com/sandeco/reversa --skill reversa-challenger
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: reversa-challenger
Source: https://github.com/sandeco/reversa/tree/main/agents/reversa-challenger
Command: npx skills add https://github.com/sandeco/reversa --skill reversa-challenger

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Early-stage project options often hide fatal assumptions and unbudgeted costs, especially in legacy systems. This Skill stress-tests every option from the ideation session before a decision is made, exposing the premises that could kill the project while they are still cheap to kill.

Core Features & Use Cases

  • Premortem Analysis: Generates 2-4 distinct failure headlines with different root causes, then asks the user which one scares them most to set severity order.
  • Per-Option Attack: For each option in options.md, identifies the single killing premise, a cheap test for it, the hidden cost (anchored to concrete legacy files or specs), and the point of no return.
  • Cross-Cutting Risk Register: Checks applicable transversal risks such as single-person dependency, regulatory obligations, third-party integrations, data migration, and authentication changes.
  • Use Case: After running the Explorer agent on a legacy modernization idea, invoke the Challenger to produce risks.md in the active ideation session, so the Arbiter agent can score options against concrete risks.

Quick Start

Ask the agent to run the Challenger premortem on the current ideation session and generate the risks.md document.

Frequently Asked Questions about reversa-challenger

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

FAQPage Schema
How do I run a premortem analysis on project options with an AI agent?

Invoke the reversa-challenger skill after the Explorer has produced options.md in the active ideation session. It generates failure headlines, attacks each option's core premise, and writes the results to risks.md in the session directory.

What does the reversa-challenger agent produce?

It produces a risks.md file containing premortem headlines, per-option analysis (killing premise, cheap test, hidden cost, point of no return), applicable transversal risks, and up to five questions that must be answered before deciding.

Which AI coding assistants support the reversa-challenger skill?

The skill declares compatibility with Claude Code, Codex, Cursor, Gemini CLI, and other agents compatible with the Agent Skills format. It is installed by the Reversa framework into .agents/skills/ or .claude/skills/.

What happens if options.md is missing when running the Challenger?

The agent stops and instructs you to run /reversa-explorer first, since the Challenger requires the options document from the active ideation session. It also exits if no active ideation session exists, pointing you to /reversa-brainstorm.

Does the Challenger agent modify project source code?

No. Its absolute rule is to write only to the session's risks.md file and the current-stage field of active-ideation.json. It never touches other project files and never produces code.