Coordinate specialized agents to complete complex tasks with persistent sessions.
npx skills add https://github.com/sakthiram/chitti --skill multi-agent
Please help me install this Agent Skill. Skill: multi-agent Source: https://github.com/sakthiram/chitti/tree/main/skills/multi-agent Command: npx skills add https://github.com/sakthiram/chitti --skill multi-agent
This Skill addresses the critical limitations of single AI agents in handling complex, long-running tasks. It overcomes context limits, prevents constant interruptions by agents, and ensures high-quality outputs through structured review cycles. Engineers can delegate multi-hour tasks without constant supervision, freeing them to focus on higher-value work.
To start a new task named 'my-feature' in your current project, first set up agents by running ./scripts/setup-agents . then initiate the task with ./scripts/task start my-feature --project .. You can check its progress using ./scripts/task status my-feature.
High-intent search queries and answers about installing and using this skill.
Coordinate autonomous agents by classifying tasks and dynamically assigning specialized agents—explore, plan, architect, dev, test, review—that maintain context across sessions. Signal-based decisions from a central PM agent adapt to real-time progress, eliminating hardcoded timeouts and ensuring continuous task completion without interruption.
Yes. Agents operate on remote codebases via SSH with rsync for file synchronization and tmux for persistent sessions. This enables seamless development workflows where agents maintain context and coordinate work across distributed environments.
Run `./scripts/setup-agents .` to initialize agents, then `./scripts/task start <task-name> --project .` to begin. Track progress with `./scripts/task status <task-name>`, which logs state to pm_state.json and timestamped handoffs between agent roles.
Persistent sessions preserve agent context across interruptions. Task orchestration automatically routes work to specialized agents—each with focused scope—and the PM agent makes signal-based decisions to maintain progress without restarting from scratch or requiring manual intervention.
Yes. The skill integrates with both Claude and Kiro CLI through the claude-cli dependency, enabling agents to execute commands, maintain session state, and coordinate autonomously across multi-hour tasks.
Mandatory review cycles are enforced between agent handoffs. A dedicated review agent evaluates outputs against functional and technical requirements before downstream agents proceed, ensuring iterative feedback and high-quality deliverables.
Khamel83Execute remote commands over SSH for git sync, file transfer, and tmux jobs.
vrcmsOrchestrate scheduled autonomous agent workflows with persistent memory and MCP tool execution.
haJ1tStore and load AI agent decisions, mistakes, and context across sessions.
clineInteract with local operating systems and remote SSH machines via persistent shell sessions.
