skills.rest
snewhouse
snewhouseCOMMUNITY
Explore all skills by this author →
𝕏@s_j_newhouseGitHub

plan-verification

Verify AA-MA plans by auditing claims, assumptions, and acceptance criteria.

6|1|Updated Apr 5, 2026
Search Tags:#multi-agent#acceptance criteria#impact analysis#adversarial review#plan verification#assumption audit
One-click install
npx skills add https://github.com/snewhouse/aa-ma-forge --skill plan-verification-snewhouse
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: plan-verification
Source: https://github.com/snewhouse/aa-ma-forge/tree/main/claude-code/skills/plan-verification
Command: npx skills add https://github.com/snewhouse/aa-ma-forge --skill plan-verification-snewhouse

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill prevents wasted cycles and late-stage failures by adversarially verifying AA-MA plans against external reality before they proceed to execution.

Core Features & Use Cases

  • Six-angle, adversarial verification: evaluates factual correctness, challenged assumptions, ripple impacts, and acceptance-criteria falsifiability across multiple independent perspectives.
  • Automated or interactive modes: blocks artifact creation on CRITICAL issues in automated mode, while report-only behavior is used in interactive mode.
  • Standalone plan verification: supports re-running verification after plan revisions and via a manual /verify-plan command.

Quick Start

Use the /verify-plan command to verify a specific plan in automated or interactive mode before you execute milestones.

Frequently Asked Questions about plan-verification

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

FAQPage Schema
How do I catch errors in a multi-step plan before execution?▼

Plan verification audits factual claims, challenges assumptions, and stress-tests acceptance criteria across multiple angles to catch implementation barriers before execution. It dispatches parallel sub-agents to evaluate ripple impacts and domain risks.

What is adversarial review for multi-agent planning scenarios?▼

Adversarial review applies six independent verification perspectives to extract structured plan claims and consolidate findings by severity. It evaluates factual correctness and falsifiability to report PASS, FAIL, or PASS WITH WARNINGS.

How do I verify plan assumptions and analyze ripple effects?▼

To verify assumptions and analyze ripple effects, extract structured claims from your plan and dispatch parallel verification sub-agents. These sub-agents audit implicit assumptions and evaluate downstream impacts to consolidate findings by severity.

Does plan verification block artifact creation on critical issues?▼

In automated mode, plan verification blocks artifact creation when CRITICAL issues are detected. Interactive mode operates as report-only, allowing you to review PASS WITH WARNINGS findings without blocking your workflow.

Can I re-run verification after revising a long-horizon plan?▼

You can re-run plan verification after revisions using the manual command. This supports iterative refinement of long-horizon, multi-step plans by re-evaluating vague acceptance criteria and implementation barriers against external reality.

Related Skills

Explore More Software Engineering Skills→
ddobrinddobrin
8

implementation-validator

Validate code implementations against acceptance claims via adversarial review.

Community
Advanced
Alexander-TyagunovAlexander-Tyagunov
12

scrutinize

Dispatch three specialist reviewers to surface issues in changed code files.

Community
Advanced
alicicekalicicek
36

plan-phase

Orchestrate adversarial planning workflows for complex software engineering tasks.

Community
Advanced
vuxsoul-uivuxsoul-ui

agent-orchestrator

Decompose macro tasks into parallel subtasks and coordinate sub-agents via file-based communication.

Community
Advanced