Want to connect with Mars?
Join organizations building the agentic web. Get introductions, share updates, and shape the future of .agent.
Is this your company?
Claim this profile to update your info, add products, and connect with the community.
Mars is a directory and launchpad for AI agents, specifically those emerging from the ecosystem of on-chain autonomous entities. It tracks the creation and deployment of agents developed by Dogshit Labs, providing a central hub for observing agentic activity across social and financial protocols. This positions the company in the deployment and tracking layer of the agent stack, focusing on the social and economic lifecycle of agents rather than just their underlying model capabilities.
The platform is significant because it highlights the transition from monolithic AI models to fleets of small, specialized agents. By providing a clear distinction between agents that are "Created" and those that are "Launched," Mars emphasizes execution and live interaction. It is a key resource for those tracking the high-velocity development of autonomous bots that operate with their own wallets and social personas.
In the current AI cycle, the distance between an idea and a deployed, autonomous entity is shrinking. Mars, accessible via its primary domain at Mars.fun, is an engine for this contraction. It is not a traditional software-as-a-service platform but a high-velocity deployment engine and directory for AI agents. The project is an output of Dogshit Labs, a development group that leans into the irreverent, builder-centric aesthetic that has come to define the intersection of decentralized finance and autonomous compute.
The platform provides a minimalist interface dominated by two primary categories: Created and Launched. This binary view reflects the lifecycle of modern AI agents in the crypto-adjacent space. A "created" agent exists as a defined persona or a piece of code, while a "launched" agent is one that has been pushed live to a blockchain or a social network, often with its own economic incentive structure. By cataloging these entities in one place, Mars acts as a command center for an increasingly fragmented ecosystem of autonomous bots.
Dogshit Labs, the entity behind Mars, represents a specific subculture in the agentic world. While enterprise AI firms focus on compliance and controlled outputs, studios like Dogshit Labs prioritize activity and experimental autonomy. Their branding is intentionally abrasive, a signal to a community that values speed and permissionless innovation over corporate polish. This approach is reminiscent of the early web — chaotic, builder-centric, and indifferent to traditional institutional approval.
The technology behind these agents typically involves a mix of Large Language Models (LLMs) for reasoning and on-chain protocols for execution. While the specific technical architecture of every agent on Mars is not public, the pattern is clear: these are specialized agents designed for specific tasks or social interactions. They are part of a broader trend where agents are treated as iterative experiments rather than long-term infrastructure.
In the broader market, Mars competes with other agent launchpads and agentic token platforms. As the market for autonomous services matures, the distinction between tools for developers and tools for active participants is blurring. Mars occupies the middle ground, providing the visibility necessary for users to track what is being built while maintaining the infrastructure for rapid deployment.
Its primary differentiator is its focus on the "all" — the sheer volume of agents being pushed into the world. In a market where most projects try to highlight one or two flagship products, Mars highlights the scale of the collective output. It suggests a future where agents are not rare, but ubiquitous and ephemeral. The value lies not in any single agent, but in the platform that can successfully coordinate and display the entire fleet.
Currently, the site is a tracking layer. For developers, it is a way to prove deployment. For users, it is a discovery engine. As the agent stack becomes more standardized with tools like the Model Context Protocol (MCP) or standardized wallet integrations, platforms like Mars are likely to evolve from simple directories into orchestration layers. For now, it remains a testament to the speed of the current development cycle, where "launched" is a primary metric of success.
A dashboard for tracking and deploying autonomous agents.
Code for our 2026 robot, Marvin 19
Code to run Marvin 13.
Class to develop utility methods for Marvin 19.
2026 build of Prometheus
Mountaineer Area Robotics Team 2614 Wiki
Public robot code for "Manta"
Iron Maple's swerve-drive skeleton project, based on 6328's Advanced Swerve Drive Example
Port of ST vl53l0x driver for roboRIO to operate the REV Robotics 2m Distance Sensor.
Mars is hiring
You've explored Mars.
Join organizations building the agentic web.