CLEO is a production-grade task management system for Claude Code agents. It provides anti-hallucination protection and is designed for AI coding agents and solo developers. It connects to shell-based workflows and integrates with Claude agents.
git clone https://github.com/kryptobaseddev/cleo.gitCLEO is an agent-first task orchestration system designed for Claude Code agents and solo developers working on complex projects. It solves the coordination crisis of multi-agent workflows by providing hierarchical task management, persistent memory across sessions, and multi-provider support for Claude Code, Cursor, Gemini, and other AI coding platforms. The system includes six core components: TASKS for project management, LOOM for a 9-stage development lifecycle, BRAIN for semantic memory that doesn't decay, NEXUS for code intelligence, CANT for agent definition, and CONDUIT for agent-to-agent communication. CLEO reconnects agents to project context after interruptions and enables structured coordination where power is paired with memory and continuity.
[{"step":"Define the task and constraints","action":"Use the prompt template to specify the task (e.g., 'Deploy a Node.js app to AWS EC2'), error cases (e.g., 'AWS credentials expired'), and validation criteria (e.g., 'Port 3000 must respond').","tip":"Be explicit about environment requirements (e.g., 'Node.js 20+', 'Docker installed')."},{"step":"Run CLEO in a compatible environment","action":"Execute the task in a terminal with CLEO installed (e.g., `cleo run --task \"[YOUR_TASK]\"`). Ensure the environment matches the task requirements (e.g., Python for a Python task).","tip":"Use `cleo init` to set up a new project if starting from scratch. Test CLEO’s anti-hallucination checks by intentionally providing invalid inputs."},{"step":"Monitor execution and validate outputs","action":"Review CLEO’s step-by-step report for errors or warnings. Manually verify critical outputs (e.g., API responses, file contents) against expected results.","tip":"For complex tasks, break them into smaller CLEO subtasks (e.g., 'Build Docker image' → 'Run container' → 'Validate endpoints')."},{"step":"Iterate and log results","action":"Address any issues flagged by CLEO (e.g., retries, error handling). Save the final report and artifacts (e.g., logs, screenshots) to a designated directory.","tip":"Use CLEO’s logging features to track task history (e.g., `cleo log --task-id [ID]`). Share reports with team members for collaboration."},{"step":"Integrate with CI/CD pipelines","action":"Embed CLEO tasks in scripts (e.g., GitHub Actions, Jenkins) to automate workflows. Use CLEO’s CLI for programmatic control (e.g., `cleo run --task \"deploy\" --env staging`).","tip":"Test CLEO tasks in a staging environment before production. Use CLEO’s error handling to trigger rollback procedures if validation fails."}]
Managing multi-stage development tasks across interrupted work sessions
Coordinating multiple AI agents on complex codebases with dependency tracking
Maintaining semantic memory of patterns and learnings across coding sessions
Tracking task hierarchies, priorities, and completion lifecycle in agent workflows
No install command available. Check the GitHub repository for manual installation instructions.
git clone https://github.com/kryptobaseddev/cleoCopy the install command above and run it in your terminal.
Launch Claude Code, Cursor, or your preferred AI coding agent.
Use the prompt template or examples below to test the skill.
Adapt the skill to your specific use case and workflow.
Use CLEO to automate the following task: [DESCRIBE_TASK]. Ensure the agent follows the CLEO workflow: (1) Verify inputs and environment, (2) Execute with anti-hallucination checks, (3) Validate outputs, and (4) Log results. Include error handling for [SPECIFIC_ERROR_CASES].
### CLEO Task Execution Report
**Task**: Automate the deployment of a Python Flask API to a staging server.
**Step 1: Input Validation**
- Verified Python 3.11+ is installed (✅).
- Confirmed Docker is running (✅).
- Checked Git repository for uncommitted changes (✅).
- Validated `requirements.txt` exists (✅).
**Step 2: Execution with Anti-Hallucination Checks**
- Generated Dockerfile from template (verified against official Python 3.11-slim image).
- Built image `flask-api:staging-20241015` (size: 120MB, layers: 5).
- Ran container `flask-api-staging` (port 5000 mapped to host port 8080).
- Executed health check: `curl http://localhost:8080/health` → `{"status": "ok"}` (✅).
**Step 3: Output Validation**
- API endpoint `/api/users` returns expected JSON structure (✅).
- Load test simulated 100 requests: 99 successful, 1 timeout (handled by retry logic).
- Logs show no errors in `/var/log/flask-api-staging.log`.
**Step 4: Result Logging**
- Task completed in 4m 22s.
- Artifacts saved to `./cleo-output/flask-api-staging-20241015/`.
- Error handling triggered for timeout (retry succeeded).
**Recommendations**:
- Add database connection validation in future runs.
- Monitor container CPU/memory usage during peak load.AI assistant built for thoughtful, nuanced conversation
Get more done every day with Microsoft Teams – powered by AI
Automate security compliance and monitor real-time security posture seamlessly.
Automate your spreadsheet tasks with AI power
Agentic AI Workflow platform
Connected workspace for docs, wikis, and projects
Take a free 3-minute scan and get personalized AI skill recommendations.
Take free scan