concession-validator

Validate agent-proposed concessions against attempt logs and taxonomy rules.

1|Updated Mar 24, 2026
One-click install
npx skills add https://github.com/EdgeApp/edge-dev-agents --skill concession-validator
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: concession-validator
Source: https://github.com/EdgeApp/edge-dev-agents/tree/main/.cursor/skills/concession-validator
Command: npx skills add https://github.com/EdgeApp/edge-dev-agents --skill concession-validator

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires jq.

What problem does it solve?

This skill prevents autonomous agents from prematurely giving up on tasks or delivering incomplete work by enforcing a strict, evidence-based validation process for any proposed concessions.

Core Features & Use Cases

  • Adversarial Verification: Evaluates claimed blockers against structured attempt logs to distinguish between genuine technical walls and avoidable failures.
  • Downgrade Management: Ensures that any delivery made without reaching full in-app success is backed by documented workarounds or sanctioned fallbacks.
  • Use Case: When an agent attempts to mark a task as blocked due to a missing dependency or a perceived error, this skill forces the agent to prove it attempted the necessary workarounds, such as linking local branches or building test harnesses, before accepting the concession.

Quick Start

Invoke the concession validator for the current task ID and reason string to determine if the proposed block or downgrade is legitimate.

Frequently Asked Questions about concession-validator

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

FAQPage Schema
How do I validate agent task abandonment in automated workflows?

You validate agent task abandonment by cross-referencing stated failure reasons against structured attempt logs and predefined taxonomy rules. This process ensures autonomous agents cannot prematurely give up on tasks without proving they hit a genuine technical wall.

How do I prevent autonomous agents from delivering incomplete work?

Preventing incomplete work delivery requires enforcing an adversarial verification process for proposed concessions. This ensures that any delivery made without reaching full in-app success is backed by documented workarounds or sanctioned fallbacks.

What is concession validation in agent governance?

Concession validation in agent governance is an adversarial gatekeeping mechanism that evaluates claimed blockers against structured attempt logs. It distinguishes between genuine technical walls and avoidable failures to enforce high-quality delivery standards in automated development workflows.

Do I need jq to validate agent concession claims?

Yes, you need jq installed to use the concession validator. The skill requires jq as a dependency to parse local state directories containing attempt logs and execute verdict-writing logic based on specific concession categories.

How do I verify agent blockers against attempt logs?

You verify claimed blockers by invoking the concession validator for the current task ID and reason string. The validator cross-references these inputs against structured attempt logs to determine if the proposed block or downgrade is legitimate.

When should I use adversarial gatekeeping for agent downgrades?

You should use adversarial gatekeeping when an agent attempts to mark a task as blocked due to a missing dependency or perceived error. It forces the agent to prove it attempted necessary workarounds, such as linking local branches or building test harnesses, before accepting the concession.