At Modelit, we are looking for a Software Engineer to design and build the agentic interface to an AppExchange product: the MCP servers, CLI commands, APIs, agent skills, plugins and agents that let AI assistants and users interact with the product. You will work within an AI-assisted delivery framework, where your job is to set the intent, judge what comes back and own the result. Excellent English communication skills, both written and spoken, are required.

The ideal candidate should have a solid understanding of the concepts below and a desire to keep learning, with support from our career development programs, flexible culture and certification opportunities.

Responsibilities

• Build MCP servers: design the tools agents call, what they take in and return, how they report errors and how they evolve between versions.

• Extend a CLI with new commands that people and agents can run.

• Connect to Salesforce REST, Tooling, Metadata and Bulk APIs, including authentication and connected-app setup, and build custom REST APIs.

• Write skills and package them into plugins for different AI models, keeping them consistent across vendors (Anthropic, OpenAI, etc.).

• Design the subagents, hooks and commands that keep agents on the right path, and the small TypeScript runtimes they call.

• Write tests first and build evaluations that check whether a tool or skill gives the right result, and keep them running in CI.

• Publish to npm, review agent output for wrong assumptions and security gaps, and protect tools against untrusted input and prompt injection.

• Write the documentation and examples that let others use and extend what you build.

Requirements

• Hands-on use of coding agents on real work (Claude Code, Codex, Cursor or similar).

• Working knowledge of TypeScript and Node.js, enough to read, review and adjust what agents write.

• Solid engineering fundamentals: testing, error handling, security, performance, change control (GitHub) and CI/CD.

• Understanding of agent concepts: tool calling, MCP, skills, subagents, hooks and plugins.

• Exposure to APIs and integration patterns, such as REST APIs.

• Evaluation thinking: how to check that an AI feature gives the right answer, consistently.

• Fast iteration: design, build, test and validate in tight loops, without clinging to the first version.

• Judgment to tell when an answer looks right but is wrong, and to question the agent's output, your own assumptions and the brief itself.

• Ownership of outcomes, following problems through to the end.

• Curiosity, comfort with ambiguity and a team mindset: sharing your setup, giving useful feedback and turning repeated mistakes into team rules.

• Strong communication skills in English (oral and written).

Nice to Have

• Experience building a skill, plugin, MCP server, slash command or agent.

• Salesforce experience as an admin or developer.

• Published npm packages or maintained plugins or extensions others install.

• Experience writing developer documentation, tutorials or customer-facing guides.

• Knowledge of Agentforce and Salesforce AI products.

Benefits

• Paid Time Off – because rest matters.

• On-site (from our offices at Sinergia), hybrid, or remote work – your choice!

• Flexible hours – start your day anytime between 8–10am and wrap up between 5–7pm.

• Career growth opportunities – we grow together.

• Ongoing training & development – keep learning, keep leveling up.

• Certification vouchers on us – we’ve got your back when you’re ready to get certified.

• A team culture you’ll love – supportive, collaborative, and fun.