Detect AI agent stalls and route failures to Zeus for resolution.
npx skills add https://github.com/GunjanGrunge/rrq --skill escalation-gunjangrunge
Please help me install this Agent Skill. Skill: escalation Source: https://github.com/GunjanGrunge/rrq/tree/main/skills/escalation Command: npx skills add https://github.com/GunjanGrunge/rrq --skill escalation-gunjangrunge
This Skill prevents agent failures from halting progress by providing a robust protocol for handling situations where an AI agent cannot converge on a solution, ensuring that no job is silently abandoned and that user intervention is managed effectively.
Use the escalation skill to handle a failed quality gate for the video titled 'Introductory Montage'.
High-intent search queries and answers about installing and using this skill.
Agent failure handling works by detecting when retry loops yield no meaningful improvement, routing the stuck state to a centralized evaluation protocol, and triggering user notifications with defined actions and timeouts to ensure no job is silently abandoned.
Automated stuck detection identifies when an agent's retry loops are not yielding meaningful improvement, routing the stalled state to a centralized evaluation protocol to attempt a fix, suggest new approaches, or escalate to a human.
User notification during an escalation is managed by sending context and clear action options with defined timeouts and auto-decisions, ensuring human judgment is requested only when necessary while preventing permanent job abandonment.
Yes, a centralized escalation protocol handles diverse scenarios including quality gates, content detection, and deadlocks across multiple domains by routing all agent failures to a single owned protocol for consistent evaluation and resolution.
An image generation agent fails to meet quality standards when it cannot converge on a solution; the escalation protocol detects the stall, allows a centralized manager to attempt a fix, and notifies the user with options to approve, abort, or retry with a different strategy.
Oruga420Orchestrate continuous autonomous agent loops with quality gates and recovery controls.
p-sree-sai-pavanCoordinates multi-step workflows with human approval gates and resume tokens.
GratefulJinx77Automate developer workflows including CLI generation, skill scaffolding, and agent delegation.
lucking7Execute tasks directly without persona-based sub-agent dependencies.
