feat: redis, fileserver, httpclient 런타임 시점 구현 추가

CLAUDE.md

@@ -4,49 +4,46 @@ Repository guidance for the Java 21 + Spring Boot 4.0.0 Clean Architecture templ
 
 ## Prime Directive
 
-Preserve architecture before optimizing for speed. The HARD-STOP SSOT is
-`.agents/plugins/ca-superpowers/rules/clean-architecture.md`; its eight current stop conditions are:
+Preserve architecture before optimizing for speed. The following eight HARD-STOP conditions are a
+synchronized summary of the canonical local policy in `AGENTS.md`:
 
 1. `domain-core` gains framework, transport, database, or cloud dependencies.
 2. A controller directly uses a repository, Spring Data interface, or persistence entity.
 3. An inbound DTO leaks into `application-core` or `domain-core`.
 4. Business rules move into mappers, filters, configuration, settings, or controllers.
-5. Project dependencies violate `.harness/project/modules.yaml` and Gradle verification.
+5. Project dependencies violate `src/config/architecture/modules.json` or the Gradle dependency
+   gate.
 6. Completion is claimed without the relevant verification or a named reason it could not run.
-7. A corpus conclusion is made without evidence appropriate to the selected evidence profile.
+7. A repository/corpus conclusion is made without evidence proportional to its scope and risk.
 8. Non-trivial work closes without the required LLM Wiki capture or a reported capture block.
 
-If this summary and the SSOT differ, the SSOT wins.
+If this summary drifts from `AGENTS.md`, `AGENTS.md` wins and this summary must be resynchronized.
 
-## Harness policy authorities
+## Gradle policy authorities
 
-- Project manifest and stack: `.harness/manifest.yaml`
-- Leaf modules, dependency edges, and focused commands: `.harness/project/modules.yaml`
-- Risk classification: `.harness/core/risk-policy.yaml`
-- Evidence selection: `.harness/core/evidence-policy.yaml`
-- Orchestration and advisory depth: `.harness/core/review-policy.yaml`
-- Durable/concise reporting: `.harness/core/report-policy.yaml`
-- Physical parity check: `.harness/validators/validate_policy_parity.py`
-
-Resolve a task packet once and reuse its stable task-packet hash and relevant rule hashes. Rerun
-resolution or reread a full rule only when the task overlay, packet hash, or rule hash changes.
+- `src/config/architecture/modules.json`: exactly 19 leaf identities, repository-relative source
+  paths, Gradle paths, and allowed production project dependency edges.
+- `src/settings.gradle`: fail-closed registry validation, project inclusion, and directory mapping.
+- `src/build.gradle`: `verifyCleanArchitectureDependencies` and the other architecture-wide
+  verification tasks.
 
 Commit policy is `human-only`: agents do not stage, commit, amend, or push implementation changes.
 
-## Risk-based orchestration
+## Proportional workflow
 
-- Low risk: implement inline or in a focused lane and run the focused check.
-- Medium risk: use `ca-implementer`; select proportional review based on affected boundaries and
-  evidence needs.
-- High risk: use `ca-implementer`, then the full chain after a human commit:
-  `ca-architect-sentinel` → `ca-spec-reviewer` → `ca-quality-reviewer` → `gradle-runner`.
+- Low risk: work in the owning leaf, follow its nearest guidance, and run the focused check.
+- Medium risk: use the relevant Superpowers design, planning, TDD, debugging, and review workflows
+  in proportion to the affected boundaries.
+- High risk: make architecture and behavior decisions explicit, use staged architecture/spec/quality
+  review, and run architecture-wide verification authorized by the task.
 
-Risk comes from change surface and task flags, not file count.
+Risk comes from change surface and runtime, security, data, or public-contract impact, not file
+count.
 
 ## Module families
 
-`.harness/project/modules.yaml` owns the complete 19-leaf list. Root guidance summarizes families;
-the nearest `src/**/CLAUDE.md` owns local rules.
+`src/config/architecture/modules.json` owns the complete 19-leaf list. Root guidance summarizes
+families; the nearest `src/**/CLAUDE.md` owns local rules.
 
 | Family | Responsibility | Stable dependency direction |
 | --- | --- | --- |
@@ -60,7 +57,8 @@ the nearest `src/**/CLAUDE.md` owns local rules.
 | `app-bootstrap` | Spring Boot entrypoint and composition root | registered runtime leaves |
 
 Never infer an individual leaf's Gradle path, allowed dependency, or test command from this table.
-Read it from `.harness/project/modules.yaml` or from the resolved task packet.
+Read its `gradle_path` and `allowed_dependencies` from
+`src/config/architecture/modules.json`; derive the focused test from that Gradle path.
 
 ## Layer workflow
 
@@ -88,7 +86,9 @@ or writable scope, stop and request context rather than expanding silently.
 - identifier: pure deterministic unit tests.
 - bootstrap/settings: binding, validation, wiring, and architecture tests.
 
-From `src/`, run the exact focused command emitted in the task packet. Architecture-wide commands:
+From `src/`, read the owning leaf's `gradle_path` from
+`config/architecture/modules.json` and run `./gradlew <gradle-path>:test --console=plain`.
+Architecture-wide commands:
 
 ```bash
 ./gradlew verifyCleanArchitectureDependencies --console=plain
@@ -114,6 +114,6 @@ or the regulated profile. Otherwise a concise result is allowed.
 
 ## LLM Wiki capture
 
-For non-trivial implementation or workflow changes, follow
-`.agents/plugins/ca-superpowers/rules/llm-wiki-capture.md`. If the controller explicitly excludes
-wiki writes for a dispatched task, report the handoff instead of writing outside scope.
+For non-trivial implementation or workflow changes, use the exact vault path and capture sequence in
+`AGENTS.md`. If the controller explicitly excludes wiki writes for a dispatched task, report the
+handoff instead of writing outside scope.