ScopeSentry: Catch Scope Creep Before It Costs Freelancers

In the world of freelancing, scope creep is a common challenge. A project begins with a clear agreement, but clients often request additional features, revisions, or deliverables. These incremental demands can transform a profitable project into unpaid labor. Enter ScopeSentry, an AI-powered tool designed to safeguard freelancers from scope creep.

By comparing an original project agreement with client conversations, ScopeSentry identifies requests as in_scope, potentially_out_of_scope, unclear, or needing_review. Each verdict is supported by evidence from the original sources, allowing freelancers to make informed decisions before agreeing to additional work.

ScopeSentry is built using open-source AI technologies. The core application is developed with Next.js and TypeScript, leveraging Gemma as the language model through the Ollama platform. Local inference ensures that sensitive project agreements and client communications remain private, never leaving the user's machine. Zod is used for validating structured data, while deterministic checks confirm that quoted evidence actually exists within the supplied text.

This design decision acknowledges that missing information in an agreement does not automatically classify a request as out of scope; rather, it flags uncertainty for human review.

The tool's privacy-focused architecture is a key feature. By running inference locally, ScopeSentry avoids transmitting project data to external AI services. This approach empowers freelancers to maintain control over their data and AI usage, aligning with the principles of open innovation. Open-weight models like Gemma enable developers to run the tool on their own machines, allowing for greater transparency, customization, and community-driven improvements.

ScopeSentry addresses a real-world concern: the frequency and impact of extra client requests beyond the original agreement. The author seeks feedback from real freelancers to validate the tool's effectiveness. By testing ScopeSentry with realistic project briefs and client requests, users can identify any issues with the verdicts and provide valuable insights on how such a tool could better assist them in managing scope changes.

While the project is still in development and awaits confirmed user feedback, the author invites freelancers to contribute to the testing process, helping to refine and enhance the tool's utility in the freelancing community.