@@ -9,6 +9,9 @@ and this project adheres to [Semantic Versioning](https://semver.org/).
 
 ## [Unreleased]
 
+### Removed
+- **`docs/standards/` (52 files).** It duplicated the Claude rules in `~/.claude/rules/` and had drifted: 8 files were older than the live rules, and 6 existed only in the repo. The 6 repo-only files (java, csharp, django, flask, spring-boot, TOOL-OPTIMIZATION) moved to `~/.claude/rules/` with `paths:` frontmatter. `load_framework_standards()` and `load_language_standards()` in `standards/selector.py` now read `~/.claude/rules/`, and `TestRuleFilePaths` reads it too, skipping where it is not installed. (#320)
+
 ## [2.1.0] - 2026-08-07
 
 **Step 1 emits its prompt instead of executing it, and three things that were only
 
@@ -152,12 +152,11 @@ Level 2: SDLC Execution Core (9 active steps: Steps 0-8)
 |   +-- architecture/                 # generate_system_diagram.py (shared utility)
 |   +-- setup/  bin/  tools/          # env setup; Windows .bat launchers; dev utilities (release.py, sync-version.py, etc.)
 |   +-- github_operations/  github_pr_workflow/  helpers/   # GitHub + PR workflow helpers
-+-- policies/03-execution-system/failure-prevention/  # failure-kb.json only (read by hooks/pre_tool_enforcer/policies/failure_kb.py); the historical 00-auto-fix/01-sync/02-standards/testing subtrees do not exist on disk -- see docs/ for standards content instead
++-- policies/03-execution-system/failure-prevention/  # failure-kb.json only (read by hooks/pre_tool_enforcer/policies/failure_kb.py); the historical 00-auto-fix/01-sync/02-standards/testing subtrees do not exist on disk -- standards live only in ~/.claude/rules/ (no repo copy, #320)
 +-- src/mcp/                          # In-engine copy of session-mgr + bridge (session_hooks, base/)
 +-- k8s/                              # Kubernetes manifests (deployment, service, hpa, configmap, secret)
 +-- tests/                            # 45 test files (37 unit, 4 integration, 3 e2e, 1 load)
 +-- docs/                             # ALL documentation, segregated by kind -- see docs/README.md for the index
-|   +-- standards/ (52)               # numbered rules 01-46 + per-language standards; the loaded copy is ~/.claude/rules/, NOT this one
 |   +-- policies/ (46)                # pipeline policy documentation. NOT a mirror of ~/.claude/policies/ -- see the note below
 |   +-- architecture/ (17)            # ADRs, pipeline/level design, flow diagrams, orchestration prompt
 |   +-- guides/ (14)                  # getting started, deployment, testing, troubleshooting, runbooks
@@ -202,7 +201,7 @@ unversioned: if that machine is lost, so are they.
 | Integrations Package | langgraph_engine/integrations/ | Abstract Factory + Lifecycle: GitHub/Jira/Figma/Jenkins |
 | Level 0 | langgraph_engine/preflight_guard/ | Pre-Flight Sanity Guard -- auto-fix enforcement (canonical) |
 | Level 1 | langgraph_engine/context_sync/ | Session & Context Synchronization (canonical). Outputs: `complexity_score` [1-10] (simple heuristic), `combined_complexity_score` [1-25] (simple x 0.3 + graph x 0.7 after linear scaling). **Note: `combined_complexity_score` is on a 1-25 scale -- do NOT treat it as 1-10.** |
-| Standards (non-numbered) | docs/ (standards) + langgraph_engine/standards/ (selector + library_adapter) | Standards policies (.md files, no pipeline nodes) -- always-on, loaded from disk; retired from the level count since it has never had pipeline nodes |
+| Standards (non-numbered) | ~/.claude/rules/ (the only copy; no repo docs/standards/) + langgraph_engine/standards/ (selector + library_adapter) | Standards policies (.md files, no pipeline nodes) -- always-on, loaded from disk; retired from the level count since it has never had pipeline nodes |
 | Level 2 | langgraph_engine/sdlc_pipeline/subgraph.py | SDLC Execution Core -- 9-step active execution (Steps 0-8) -- ACTIVE (nodes in sdlc_pipeline/nodes/) |
 | Pre-Analysis Node | langgraph_engine/sdlc_pipeline/subgraph.py | orchestration_pre_analysis_node: CallGraph scan at Step 0; template fast-path detection |
 | Hooks | hooks/pre-tool-enforcer.py, post-tool-tracker.py, stop-notifier.py | Tool enforcement + session maintenance |
 
@@ -475,7 +475,7 @@ claude-workflow-engine/           # 369 Python files total
 │   ├── e2e/                      # 3 end-to-end scenario tests
 │   └── load/                     # 1 concurrency / load test
 │
-├── docs/                         # 253 files — architecture docs, ADRs, runbooks, ALL policy/rule
+├── docs/                         # 201 files — architecture docs, ADRs, runbooks, ALL policy/rule
 │                                 # .md files (flattened here, no subfolders), CONTRIBUTING.md,
 │                                 # CODE_OF_CONDUCT.md, and the GitHub issue/PR templates
 ├── uml/                          # Regenerated per pipeline run (13 types, Mermaid/PlantUML)
@@ -814,7 +814,7 @@ See [CHANGELOG.md](CHANGELOG.md) for the complete version history.
 
 **Complexity score is not ground truth:** `combined_complexity_score` is a heuristic: `simple_score × 0.3 + graph_score × 0.7`. It is on a 1-25 scale and correlates with effort, but it does not map to story points and should not be treated as precise.
 
-**Framework-specific standards cover only Flask, Django, and Spring Boot so far.** `load_framework_standards()` in `standards/selector.py` looks in `docs/standards/` for a `{framework}-standards.md` file; `docs/standards/flask-standards.md`, `docs/standards/django-standards.md`, and `docs/standards/spring-boot-standards.md` are bundled. `detect_framework()` also recognizes FastAPI, Pyramid, Tornado, Celery, Scrapy, plain Spring, Quarkus, Micronaut, and the common JS/TS frameworks (React, Angular, Vue, Express, etc.) — none of those have a bundled doc yet, so detection succeeds but the framework tier returns empty for them. `load_language_standards()` has full coverage: all 6 project types `detect_project_type()` recognizes except `unknown` resolve real content (`docs/standards/02-backend-standards.md` python, `docs/standards/06-typescript-standards.md` javascript/typescript, `docs/standards/07-go-standards.md` go, `docs/standards/08-rust-standards.md` rust, `docs/standards/java-standards.md` java, `docs/standards/csharp-standards.md` csharp).
+**Framework-specific standards cover only Flask, Django, and Spring Boot so far.** `load_framework_standards()` in `standards/selector.py` looks in `~/.claude/rules/` (the only copy of the Claude rules; the repo no longer keeps one) for a `{framework}-standards.md` file; `flask-standards.md`, `django-standards.md`, and `spring-boot-standards.md` exist there. `detect_framework()` also recognizes FastAPI, Pyramid, Tornado, Celery, Scrapy, plain Spring, Quarkus, Micronaut, and the common JS/TS frameworks (React, Angular, Vue, Express, etc.) — none of those have a bundled doc yet, so detection succeeds but the framework tier returns empty for them. `load_language_standards()` has full coverage: all 6 project types `detect_project_type()` recognizes except `unknown` resolve real content (all under `~/.claude/rules/`: `02-backend-standards.md` python, `06-typescript-standards.md` javascript/typescript, `07-go-standards.md` go, `08-rust-standards.md` rust, `java-standards.md` java, `csharp-standards.md` csharp). On a machine without `~/.claude/rules/` both tiers return empty and fall through to the library-skill tiers.
 
 ### Trade-offs by Design
 
 
@@ -22,70 +22,10 @@ permits: `README.md`, `CLAUDE.md`, `SRS.md`, `CHANGELOG.md` and `VERSION`.
 | [`policies/`](policies/) | Pipeline policies | 44 |
 | [`releases/`](releases/) | Per-release design notes | 7 |
 | [`reports/`](reports/) | Investigations and audits | 26 |
-| [`standards/`](standards/) | Coding and testing standards | 52 |
 
----
-
-## `standards/` -- Coding and testing standards
-
-Numbered rules 01-46 plus per-language standards. These mirror `~/.claude/rules/`, which is what the Standards mechanism loads at runtime; the copies here are for reading, not execution.
-
-<details><summary>52 file(s)</summary>
-
-- [01-common-standards.md](standards/01-common-standards.md)
-- [02-backend-standards.md](standards/02-backend-standards.md)
-- [03-microservices-standards.md](standards/03-microservices-standards.md)
-- [04-frontend-standards.md](standards/04-frontend-standards.md)
-- [05-security-standards.md](standards/05-security-standards.md)
-- [06-typescript-standards.md](standards/06-typescript-standards.md)
-- [07-go-standards.md](standards/07-go-standards.md)
-- [08-rust-standards.md](standards/08-rust-standards.md)
-- [09-swift-standards.md](standards/09-swift-standards.md)
-- [10-kotlin-standards.md](standards/10-kotlin-standards.md)
-- [11-documentation-files.md](standards/11-documentation-files.md)
-- [12-docstrings-only.md](standards/12-docstrings-only.md)
-- [13-spring-cloud-infrastructure.md](standards/13-spring-cloud-infrastructure.md)
-- [14-entity-design-patterns.md](standards/14-entity-design-patterns.md)
-- [15-dto-form-separation.md](standards/15-dto-form-separation.md)
-- [16-validation-sequence-pattern.md](standards/16-validation-sequence-pattern.md)
-- [17-api-response-wrapper.md](standards/17-api-response-wrapper.md)
-- [18-service-layer-conventions.md](standards/18-service-layer-conventions.md)
-- [19-exception-handling-hierarchy.md](standards/19-exception-handling-hierarchy.md)
-- [20-inter-service-communication.md](standards/20-inter-service-communication.md)
-- [21-caching-strategy.md](standards/21-caching-strategy.md)
-- [22-common-library-design.md](standards/22-common-library-design.md)
-- [23-enum-as-domain-model.md](standards/23-enum-as-domain-model.md)
-- [24-constants-organization.md](standards/24-constants-organization.md)
-- [25-jpa-auditing-pattern.md](standards/25-jpa-auditing-pattern.md)
-- [26-openapi-documentation.md](standards/26-openapi-documentation.md)
-- [27-centralized-logging.md](standards/27-centralized-logging.md)
-- [28-test-coverage-enforcement.md](standards/28-test-coverage-enforcement.md)
-- [29-container-deployment.md](standards/29-container-deployment.md)
-- [30-maven-build-conventions.md](standards/30-maven-build-conventions.md)
-- [31-security-authentication.md](standards/31-security-authentication.md)
-- [32-repository-conventions.md](standards/32-repository-conventions.md)
-- [33-test-case-roadmap.md](standards/33-test-case-roadmap.md)
-- [34-frontend-package-structure.md](standards/34-frontend-package-structure.md)
-- [35-positive-testing-standards.md](standards/35-positive-testing-standards.md)
-- [36-negative-testing-standards.md](standards/36-negative-testing-standards.md)
-- [37-edge-case-testing-standards.md](standards/37-edge-case-testing-standards.md)
-- [38-test-mocking-strategy.md](standards/38-test-mocking-strategy.md)
-- [39-cross-cutting-test-patterns.md](standards/39-cross-cutting-test-patterns.md)
-- [40-universal-test-patterns-abstract.md](standards/40-universal-test-patterns-abstract.md)
-- [41-testing-standards-python-fastapi.md](standards/41-testing-standards-python-fastapi.md)
-- [42-testing-standards-nodejs-express.md](standards/42-testing-standards-nodejs-express.md)
-- [43-testing-standards-go.md](standards/43-testing-standards-go.md)
-- [44-srs-lifecycle.md](standards/44-srs-lifecycle.md)
-- [45-uml-diagram-lifecycle.md](standards/45-uml-diagram-lifecycle.md)
-- [46-architecture-documentation.md](standards/46-architecture-documentation.md)
-- [TOOL-OPTIMIZATION-LEVEL2-STANDARD.md](standards/TOOL-OPTIMIZATION-LEVEL2-STANDARD.md)
-- [csharp-standards.md](standards/csharp-standards.md)
-- [django-standards.md](standards/django-standards.md)
-- [flask-standards.md](standards/flask-standards.md)
-- [java-standards.md](standards/java-standards.md)
-- [spring-boot-standards.md](standards/spring-boot-standards.md)
+Coding and testing standards (numbered rules 01-46 plus per-language standards) are not kept in this repository. They live only in `~/.claude/rules/`, which is what Claude Code and `standards/selector.py` load (#320).
 
-</details>
+---
 
 ## `policies/` -- Pipeline policies
 
 
@@ -204,7 +204,7 @@ not recomputed here and is owned elsewhere (PRD NFR-4).
 that a blocking policy returns exit code 2 from the `PreToolUse` hook so the tool call does not
 proceed. Deleting `PreToolUse` and `PostToolUse` falsifies this for two of the four events.
 
-Per `rules/44` (mirrored in this repo at `docs/standards/44-srs-lifecycle.md`) the SRS is
+Per `rules/44` (`~/.claude/rules/44-srs-lifecycle.md`; the repo copy was removed in #320) the SRS is
 append-only, so FR-9 cannot be quietly edited in place. The
 required handling is a superseding append that retires FR-9's four-event guarantee, states the v2.0.0
 replacement guarantee, and adds a Change Log row. An agent that silently rewrites FR-9 in place is in
 
@@ -42,7 +42,7 @@ actually enforced at runtime and which are prose only.
 
 Corpus to read in full:
 - `docs/policies/` — 46 markdown policy documents (the authored corpus)
-- `docs/standards/` — 52 files: numbered rules 01–46 plus per-language standards
+- `~/.claude/rules/` — numbered rules 01–46 plus per-language standards (the only copy; no repo `docs/standards/` since #320)
 - `~/.claude/policies/` — 5 directories: `01-sync-system`, `02-standards-system`,
   `03-execution-system`, `failure-prevention`, `testing` (this is what `get_policies_dir()` reads today)
 - `policies/` in-repo — contains **only** `03-execution-system/failure-prevention/failure-kb.json`
@@ -1716,7 +1716,7 @@ AGREED CONTRACTS
 
 INPUT CONTEXT
 - `structural_inventory.json`, `ast_call_graph.json`, `audit_surface.json`.
-- `docs/policies/` — 46 files. `docs/standards/` — 52 files (numbered rules 01–46 + language standards).
+- `docs/policies/` — 46 files. `~/.claude/rules/` — numbered rules 01–46 + language standards (no repo copy).
 - Known enforcement readers to check: `hooks/pre_tool_enforcer/policies/failure_kb.py` reads
   `policies/03-execution-system/failure-prevention/failure-kb.json`; `get_policies_dir()` resolves
   `~/.claude/policies/`.