Clawdbot / Moltbot
Pour une evaluation pratique : Clawdbot and Moltbot are historical names tied to the OpenClaw personal AI assistant ecosystem, useful for users searching older tutorials, repos, and community posts.
300
OpenClaw
Pour une evaluation pratique : OpenClaw is a viral open-source personal AI assistant that runs as a self-hosted, always-on agent across chat apps, local tools, memory, skills, and automation workflows.
3 views0 likes0 uses
Capabilities
- Pour une evaluation pratique : Self-hosted personal AI assistant with chat-app control through WhatsApp, Telegram, Discord, Signal, and related channels.
- Pour une evaluation pratique : Persistent memory, persona onboarding, heartbeats, background tasks, and skill-based extension patterns.
- Pour une evaluation pratique : Can coordinate coding sessions, browser actions, shell commands, calendar/email tasks, file workflows, and custom automations when permissions are granted.
- Pour une evaluation pratique : Open-source repository and official website give users a path to inspect, fork, and self-host instead of relying on a closed SaaS assistant.
- Pour une evaluation pratique : Best evaluated as a personal operating layer rather than a simple chatbot or single-purpose coding assistant.
- Pour une evaluation pratique : Requires careful security review because powerful local access, skill installation, and external integrations can expose sensitive data.
Use Cases
- Scenario d'essai: Pour une evaluation pratique : Set up a low-risk OpenClaw instance in a VM and ask it to prepare a daily brief from public RSS feeds, calendar test data, and a throwaway inbox.
- Scenario d'essai: Pour une evaluation pratique : Connect it to a demo Telegram or Discord account and test whether it can remember preferences, schedule reminders, and report progress without leaking private data.
- Scenario d'essai: Pour une evaluation pratique : Let it create a small custom skill for a harmless workflow, such as summarizing public GitHub issues or checking a test website uptime.
- Scenario d'essai: Pour une evaluation pratique : Evaluate coding-agent handoff by asking it to trigger Codex or Claude Code on a toy repository, run tests, and summarize the diff.
- Scenario d'essai: Pour une evaluation pratique : Trial family or team assistant workflows only after separating credentials, limiting file access, and reviewing every installed skill.
- Scenario d'essai: Pour une evaluation pratique : Compare with Claude Code, Codex CLI, Manus, Dify, CrewAI, and LangGraph depending on whether you need personal automation, software engineering, or agent app building.
Examples and Source Notes
- Site officiel: https://openclaw.ai/
- Depot officiel: https://github.com/openclaw/openclaw
- Documentation: https://docs.openclaw.ai/
- Source logo/image: GitHub organization avatar uploaded as the visible agent logo; official site and repo are cited in examples.
- Controle des risques: Do not run it on a primary workstation with unrestricted credentials during evaluation; inspect skills, OAuth scopes, browser permissions, and exposed admin surfaces first.
Related Agents
personal-agent