Impasse Detector

Detect agent reasoning loops from transcripts and tool outputs.

27|8|Updated Jan 16, 2026

Search Tags:#automation #safety #PowerShell #transcripts #detect #impasse #loop-detection

Ready to run

npx skills add https://github.com/Sounder25/Google-Antigravity-Skills-Library --skill impasse-detector

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: Impasse Detector
Source: https://github.com/Sounder25/Google-Antigravity-Skills-Library/tree/main/17_impasse_detector
Command: npx skills add https://github.com/Sounder25/Google-Antigravity-Skills-Library --skill impasse-detector

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill identifies when an agent gets stuck in a reasoning loop or unproductive state by examining transcripts, tool usage, and sentiment patterns.

Core Features & Use Cases

Automated loop detection: Monitors dialogue and actions to surface impasses and escalate or delegate appropriately.
Contextual analysis: Analyzes recent history using a lookback window to determine stagnation.
Use Case: In long-running planning sessions, trigger a review when looping behavior is detected to avoid wasted tokens and time.

Quick Start

To activate, provide either a transcript path or direct content and a lookback window; the SKILL will output a JSON with status, confidence, reasons, and recommended action.

Frequently Asked Questions about Impasse Detector

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

FAQPage Schema

How do I detect when an agent is stuck in a reasoning loop?▼

Detect agent reasoning loops by analyzing transcripts and tool outputs within a lookback window to surface stagnation. This process identifies repetitive dialogue and action patterns, outputting a structured JSON status with confidence and recommendation fields.

What inputs are needed to analyze transcripts for agent stagnation?▼

Analyzing transcripts for agent stagnation requires a transcript path or direct content and a lookback window. Supplying these inputs allows the tool to evaluate recent history and output a structured JSON with status, confidence, reasons, and recommended action.

Can I use PowerShell scripts to monitor long-running planning sessions for impasses?▼

Yes, you can use PowerShell scripts to monitor long-running planning sessions for impasses. By processing transcripts and tool usage patterns, the script detects unproductive states and triggers escalation or delegation to avoid wasted tokens and time.

What is automated loop detection and how does it apply to ongoing planning sessions?▼

Automated loop detection is the process of monitoring dialogue and actions to surface impasses in ongoing planning sessions. It analyzes recent history using a lookback window to determine stagnation and triggers escalation or delegation to avoid wasted tokens.

Are there limitations to using transcript analysis for detecting unproductive agent states?▼

Limitations of using transcript analysis for detecting unproductive agent states include reliance on a defined lookback window and accurate transcript data. Without sufficient recent history or direct content, the tool cannot effectively surface stagnation or recommend appropriate delegation.