A business analyst built a Copilot triage agent that labels incoming issues and drafts acceptance criteria for a mobile checkout project with 40 stakeholders. The team needs the agent discoverable inside that GitHub project for reuse. Which approach should the analyst use to save an agent to a GitHub project?

The correct approach works because GitHub project-scoped agents inherit the project's membership and context, so saving the triage agent to the project workspace makes labeling logic and acceptance-criteria prompts immediately discoverable and reusable by all 40 mobile checkout stakeholders without extra sharing steps, preserving issue context across board views and workflows. Publishing to the organization library fails by scoping concept because enterprise-wide discovery dilutes ownership and does not anchor the agent inside the specific project where checkout issues are triaged. Saving to the linked repository fails by container concept because repository-level assistants serve code-centric tasks like pull requests and file search, not cross-issue project coordination, so team members working from the project board would not reliably find it. Saving to the personal library fails by sharing concept because personal storage supports individual productivity across assignments but provides no project-level visibility for collaborative triage. As an exam caveat, distinguish carefully among personal, repository, project, and organization scopes, since stems often test whether reuse must be private, code-local, board-local, or enterprise-wide. As a practical check, after saving open the project as a second project member and confirm the agent appears in the project's agent picker and correctly labels a test issue and drafts criteria before announcing it to stakeholders.