원클릭으로
outside-in-tdd
RGR sequence, observed-failure evidence, drill-down unit tests, and non-behavioral exemptions for Phoenix/Commanded work.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
메뉴
RGR sequence, observed-failure evidence, drill-down unit tests, and non-behavioral exemptions for Phoenix/Commanded work.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
| name | outside-in-tdd |
| description | RGR sequence, observed-failure evidence, drill-down unit tests, and non-behavioral exemptions for Phoenix/Commanded work. |
Use this skill for behavior changes and bug fixes. This skill defines the discipline; the specialist RGR agents perform the writing and review handoffs. Prefer the full outside-in-rgr-microcycle workflow whenever code will be written.
Never write production behavior without an observed failing test demanding it.
rgr-test-author to write or activate that test, run the focused command, and capture real failing output.rgr-test-reviewer to approve the RED evidence and API pressure before production edits.rgr-diagnostic-implementer to implement only the minimum code that changes one current diagnostic.rgr-implementation-reviewer to approve the GREEN diff before refactor or broader verification.When an integration or acceptance failure points at internal logic, route the lower-level unit test through rgr-test-author and rgr-test-reviewer, observe it fail, use rgr-diagnostic-implementer for the minimum GREEN change, then return to the outer test.
Observed failure output must be copied from an actual run, not paraphrased. Commit bodies should explain why and include the RED command/output for behavior commits when practical.
RED is not required for docs-only changes, pure renames or moves where existing tests cover behavior, generated lockfile updates, and mechanical config chores. If a production Elixir/Phoenix/Commanded edit changes observable behavior, the exemption does not apply.
Fine-grained outside-in RED-GREEN-REFACTOR microcycles with specialist agents and single-diagnostic implementation.
Phoenix 1.8, Postgres, Commanded, Ecto, ExUnit, Dialyzer, and strict static-check conventions.
Phoenix auth/session/CSRF, command authorization, event data sensitivity, Postgres, dependency, secret, and threat-model review triggers.
Process Forgejo PR feedback with reflection, guardrail-gap classification, and inline thread replies.
Classify review feedback and choose proportionate guardrails without inflating default context cost.
Write implementation plans as test-addressed red-green-refactor cycles rather than component waterfalls.