Crewloom
Apache-2.0 toolkit with 47 specialist AI agent roles, working memory, MCP integration, and isolated Docker execution for existing agent hosts.
At a Glance
Apache-2.0 licensed public toolkit. Model-provider usage, infrastructure, and separately marketed proprietary engine are not included.
Engagement
Available On
Alternatives
Listed Nov 2026
About Crewloom
Crewloom is a bilingual English/Arabic AI-agent toolkit from RUMUZE Agency. The current public repository provides 47 specialist roles, working memory, registered tools, an MCP bridge, and isolated Docker execution. It is licensed Apache-2.0.
Public toolkit
The repository supplies roles and playbooks for existing agent hosts, project-scoped context and memory, a CLI, an optional local dashboard, and managed execution workflows. Git and Python 3.9 or newer are required for the core; isolated workflows also require Docker. The repository states that it does not include a hosted orchestration server or a model provider and is distributed through Git rather than a published npm or PyPI package.
Scope and website differences
The marketing website separately advertises desktop beta v0.5.10 and a proprietary agent engine, as well as Telegram/WhatsApp bridge and local-model capabilities. These claims should not be treated as the license or packaging of every public repository component. The public toolkit license does not establish free access to a separate proprietary engine, hosted service, or model provider.
Verification
Roles guide the host agent, and repository checks validate structure and recorded evidence. Marketing claims such as zero destructive regressions and sub-millisecond execution are not general guarantees of safe behavior or independent performance measurements. Teams must verify changes and supply their own model access and project inputs.
Community Discussions
Be the first to start a conversation about Crewloom
Share your experience with Crewloom, ask questions, or help others learn from your insights.
Pricing
Public toolkit — Apache-2.0
Apache-2.0 licensed public toolkit. Model-provider usage, infrastructure, and separately marketed proprietary engine are not included.
- 47 specialist roles
- Working memory and MCP bridge
- Isolated Docker workflows
- Optional local dashboard
Capabilities
Key Features
- 47 specialist AI-agent roles
- English and Arabic workflows
- Project-scoped working memory and context
- MCP bridge for agent hosts
- Isolated Docker execution
- Optional local dashboard
- Concurrent task worktrees
