Agent Harness Engineer

Manus AI; Singapore; On-site; SGD 110,000 - 140,000; Full time

Benefits offered by this job

Equity incentive

Unlimited Manus Tokens

Job summary

Manus AI is seeking an experienced engineer to build core systems for the agent runtime and control layer, including context, memory, tools orchestration and safety boundaries. You will own complex components and drive design, rollout and improvements for reliable performance on real tasks.

You will collaborate across product, research, infra and security to review designs, identify risks and ensure high engineering quality, with a focus on latency, cost and observability in a fast-growing AI

Qualifications

• 3–5 years of software engineering experience or equivalent ability.
• Backend or distributed systems experience covering concurrency, task queues, persistence, idempotency, failure recovery and observability.
• Ability to own complex systems or critical components independently, exercise technical judgment and deliver quality outcomes.
• Effective use of AI for research, coding and validation while retaining independent judgment.
• Ability to collaborate on technical work in Mandarin and to read, write and participate in day-to-day technical communication in English.
• Production experience with agent runtimes, harnesses, multi-step workflows, task scheduling or code execution systems is preferred.
• Deep experience in context engineering, long-term memory, retrieval and ranking, model routing, tool use, MCP, multi-agent systems or agent evaluation is preferred.
• Experience with execution isolation, least privilege, multi-tenancy, prompt-injection defenses, high-risk action approvals or auditing is preferred, as is open-source work or technical writing.

Responsibilities

• Own core components for context and memory, model and tool orchestration, task planning and execution, tools and MCP integrations, and evaluation and quality feedback.
• Define problems and success criteria when requirements and technical approaches are unclear, and independently drive technical decisions, design, implementation, rollout and ongoing improvement.
• Improve agent performance through context engineering, prompt design, retrieval, tools and architecture, evaluating task success, reliability, latency and cost.
• Build reproducible agent evaluations, regression tests and end-to-end observability, using real tasks and production failures to guide system improvements.
• Work with product, research, infrastructure and security teams on design reviews, risk identification and delivery, and improve engineering quality.

Skills

Software engineering

Distributed systems

Concurrency

Observability

English communication

Mandarin collaboration

Job description

About the Role

Build core systems for the agent runtime and control layer, covering context, tools, execution, feedback and safety boundaries. Own the design and delivery of complex components to improve reliability on real tasks.

Responsibilities

• Own core components for context and memory, model and tool orchestration, task planning and execution, tools and MCP integrations, and evaluation and quality feedback.
• Define problems and success criteria when requirements and technical approaches are unclear, and independently drive technical decisions, design, implementation, rollout and ongoing improvement.
• Improve agent performance through context engineering, prompt design, retrieval, tools and architecture, evaluating task success, reliability, latency and cost.
• Build reproducible agent evaluations, regression tests and end-to-end observability, using real tasks and production failures to guide system improvements.
• Work with product, research, infrastructure and security teams on design reviews, risk identification and delivery, and improve engineering quality.

Requirements

• Strong software engineering and computer systems foundations, with 3-5 years of software engineering experience or equivalent ability, and the ability to deliver maintainable, testable production systems.
• Backend or distributed systems experience covering concurrency, task queues, persistence, idempotency, failure recovery and observability.
• Ability to own complex systems or critical components independently, exercise technical judgment, control complexity and take responsibility for delivery quality and outcomes.
• Effective use of AI for research, coding and validation while retaining independent judgment, with initiative, clear communication and dependable collaboration.
• Ability to collaborate on technical work in Mandarin and to read, write and participate in day-to-day technical communication in English.
• Production experience with agent runtimes, harnesses, multi-step workflows, task scheduling, browser automation, CI runners or code execution systems is preferred.
• Deep experience in context engineering, long-term memory, retrieval and ranking, model routing, tool use, MCP, multi-agent systems or agent evaluation is preferred.
• Experience with execution isolation, least privilege, multi-tenancy, prompt-injection defenses, high-risk action approvals or auditing is preferred, as is open-source work, technical writing or equivalent proprietary engineering work.

Manus excels at various tasks in work and life, getting everything done while you rest at Manus AI.

What we offer:

• Build at the Frontier of AI Agents - Work with a fast-paced team to turn frontier AI breakthroughs into real-world impact.
• Equity & Shared success - Our share incentive plan enables employees to participate in Manus’s long-term growth and share in the value we create together.
• Unlimited Manus Tokens - Enjoy unlimited Manus tokens to experiment, build, and supercharge your productivity.

If you're passionate about cutting-edge technology and making a real impact, we’d love to hear from you!