parallel-agents

Orchestrates multiple specialized agents for multi-domain code analysis and review.

Updated Apr 4, 2026
One-click install
npx skills add https://github.com/juanjo-zurich/juarvis-v4 --skill parallel-agents-juanjo-zurich
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: parallel-agents
Source: https://github.com/juanjo-zurich/juarvis-v4/tree/main/plugins/core/skills/parallel-agents
Command: npx skills add https://github.com/juanjo-zurich/juarvis-v4 --skill parallel-agents-juanjo-zurich

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Complex tasks like comprehensive code audits or full-stack feature reviews require expertise across security, backend, frontend, testing, and more. Coordinating these perspectives manually is slow and error-prone. This Skill provides structured patterns for invoking and synthesizing results from multiple specialized agents within a single session. ## Core Features & Use Cases - Orchestration Patterns: Predefined workflows for comprehensive analysis, feature review, and security audits using 17 specialized agents (security-auditor, backend-specialist, test-engineer, and more). - Resilience Mechanisms: Configurable retry with exponential or linear backoff, fallback agent matrices, and shared context persisted to .juar/contexts/ for state tracking across steps. - Synthesis Protocol: A standardized report format that consolidates findings from all agents into prioritized recommendations and action items. - Use Case: Run a full security audit by chaining security-auditor and penetration-tester, then synthesize their findings into a single prioritized remediation plan. ## Quick Start Ask the AI to use the security-auditor and backend-specialist agents to review the API endpoints and synthesize the findings into one report.

Frequently Asked Questions about parallel-agents

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

FAQPage Schema
How do I run multiple AI agents on the same codebase?▼

Invoke agents sequentially with natural language commands, such as asking the explorer-agent to map the structure first, then domain specialists to analyze their areas. Findings are passed between agents and merged into a single synthesis report.

What agents are available for code review and security audits?▼

Seventeen specialized agents are available, including security-auditor, penetration-tester, backend-specialist, frontend-specialist, test-engineer, and database-architect. Each is triggered by domain keywords like "security", "API", or "tests".

When should I not use multi-agent orchestration?▼

Avoid orchestration for simple single-domain tasks, quick fixes, or small changes where one agent suffices. The overhead of coordinating multiple agents only pays off for complex, multi-perspective work.

What happens when an agent fails or times out during orchestration?▼

The retry configuration reruns the agent up to three times with exponential or linear backoff for transient errors like timeouts and rate limits. If retries exhaust, a fallback agent such as test-driven can take over based on the fallback matrix.

Can I resume a previous agent's work in a later session?▼

Yes, you can resume a previous agent by referencing its agentId and providing additional requirements. Shared context is persisted to .juar/contexts/ as YAML files with a 24-hour cleanup timeout.