Standalone Requirements

Standalone requirements capture traceable behavior that is not owned by a larger epic. They use the STANDALONE-NNN namespace and one Markdown file per requirement: NNN.md (for example, 001.md contains STANDALONE-001). Scenario IDs use the requirement ID and a stable scenario suffix, such as STANDALONE-001-S01.

Each requirement is an agent-executable specification and a human-readable feature description. Use the existing project conventions: explicit status, purpose/scope, Given/When/Then scenarios, scenario-level required evidence or validation, exclusions where useful, and an implementation record updated as work proceeds. Tests must include the exact scenario ID in the test title or native metadata, and the requirement's evidence record must link to those tests. Do not renumber committed IDs; add scenarios with the next unused suffix. Keep implementation tasks and their status in this record when the feature is small enough for one task, or link focused task records when it needs several.

Standalone requirements do not silently change an epic's scope. Where a requirement adds behavior previously excluded by an epic, record that boundary here and update epic documentation only if the product owner explicitly brings it into that epic.

Scenario-to-evidence index