HumanBean17 · HumanBean17 · Jun 7, 2026 · Jun 7, 2026
@@ -212,5 +212,5 @@ The default embedding model is `sentence-transformers/all-MiniLM-L6-v2` (downloa
 
 - `get_service_topology` — microservice-level summary aggregating `HTTP_CALLS` / `ASYNC_CALLS`.
 - Agentic routing layer (query classifier → vector / graph / both).
-- Incremental Kuzu updates (per-changed-file) — see [`propose/TIER2-INCREMENTAL-REBUILD-PROPOSE.md`](./propose/TIER2-INCREMENTAL-REBUILD-PROPOSE.md) and [`propose/INDEX-AUTO-MODE-PROPOSE.md`](./propose/INDEX-AUTO-MODE-PROPOSE.md).
+- Incremental Kuzu updates (per-changed-file).
 - Optional `codegraph_nodes` LanceDB table embedding symbol summaries so the graph itself is vector-searchable.
@@ -32,7 +32,7 @@ Notes:
 
 - `--proposal` paths may be absolute or relative to `--proposal-dir`
 - when `--proposal` is provided, `--glob` is ignored
-- add `--include-completed` if selected files can come from `propose/completed/`
+- add `--include-completed` if selected files can come from `propose/completed/` or `propose/stale/`
 
 ## Generate a propose-only workflow bundle
 
@@ -109,7 +109,7 @@ use `autopilot.py`.
   --repo-root . \
   --proposal-dir propose \
   --output-dir .agents/reports/propose_automation_selected \
-  --proposal TIER2-INCREMENTAL-REBUILD-PROPOSE.md \
+  --proposal ENHANCED-ROLE-RECOGNITION-PROPOSE.md \
   --planning-rounds 2 \
   --planning-min-severity medium \
   --implementation-rounds 2 \

@@ -208,9 +208,9 @@ disagreement as evidence of staleness, not as a contradiction.
 
 ```markdown
 **Staleness rule:** after `java-codebase-rag increment`, Lance is fresh
-but Kuzu may be stale (see `propose/TIER2-INCREMENTAL-REBUILD-PROPOSE.md`).
-A graph older than the source tree is normal mid-development. When in
-doubt, run `meta` and compare against your working tree.
+but Kuzu may be stale. A graph older than the source tree is normal
+mid-development. When in doubt, run `meta` and compare against your
+working tree.
 ```
 
 ### Patch C — `neighbors` subsection after Batching

@@ -110,8 +110,7 @@ Landing order: **PR-AGP-1 → PR-AGP-2**.
   (disagreement vs open file; empty `search` with visible string in editor).
 - Leave `After two failed attempts on the same intent, stop and report tool name, args, and response.` **unchanged**.
 - Immediately **after** that line, append the **Staleness rule** paragraph
-  (Lance vs Kuzu after `increment`, pointer to
-  `propose/TIER2-INCREMENTAL-REBUILD-PROPOSE.md`, `meta` comparison).
+  (Lance vs Kuzu after `increment`, `meta` comparison).
 
 **Patch C — `neighbors` confidence**
 

@@ -4,9 +4,8 @@ Status: **completed — shipped** to `master`. This plan implemented
 [`propose/completed/CLI-SCENARIOS-PROPOSE.md`](../propose/completed/CLI-SCENARIOS-PROPOSE.md).
 **PR-CLI-1** (propose approval + plan tracking) was [#72](https://github.com/HumanBean17/java-codebase-rag/pull/72).
 
-Depends on: **none** (engine work for Kuzu incremental rebuild stays under
-[`propose/TIER2-INCREMENTAL-REBUILD-PROPOSE.md`](../propose/TIER2-INCREMENTAL-REBUILD-PROPOSE.md)
-and the tracking issue created in PR-CLI-2). **Recommended ordering:** any
+Depends on: **none** (engine work for Kuzu incremental rebuild is tracked under
+GitHub issue #73, created in PR-CLI-2). **Recommended ordering:** any
 standalone `cli.py` import-latency / deferred-import hardening PR should land
 before PR-CLI-2 so contributors exercising new subcommands do not pay multi-second
 `--help` cost (see propose §6 closing note).
@@ -201,8 +200,8 @@ Landing order: **PR-CLI-1 → PR-CLI-2 → PR-CLI-3**.
 
 ### 8. GitHub
 
-- Open issue **“AST graph (Kuzu) incremental rebuild”** referencing
-  `propose/TIER2-INCREMENTAL-REBUILD-PROPOSE.md`; paste URL into cli constant.
+- Open issue **”AST graph (Kuzu) incremental rebuild”** (tracking issue #73);
+  paste URL into cli constant.
 
 ## Tests for PR-CLI-2
 
@@ -283,9 +282,7 @@ Follow the **explicit file list** in propose §6 (`README.md`,
 `docs/MANUAL-VERIFICATION-CHECKLIST.md`, `docs/paper/paper.tex` + rebuild
 `paper.pdf`, `AGENTS.md`, **`.cursor/rules/*.mdc`** (agent rules audit),
 `CODEBASE_REQUIREMENTS.md`, `mcp.json.example` (comments only — keys from PR-CLI-2),
-`propose/INDEX-AUTO-MODE-PROPOSE.md`,
-`propose/TIER2-INCREMENTAL-REBUILD-PROPOSE.md`, `docs/PRODUCT-VISION.md`,
-`.gitignore`).
+`docs/PRODUCT-VISION.md`, `.gitignore`).
 
 Add **Migration from legacy names** sections with explicit `mv` commands
 (`lancedb_data` → `.java-codebase-rag`, `.lancedb-mcp.yml` →

@@ -360,7 +360,7 @@ with every hit accounted for (fixed or justified).
 - Multi-target Client/Producer nodes; materialized composite edges.
 - Hints v3 implementation (PR-D — [`plans/completed/PLAN-HINTS-V3.md`](./PLAN-HINTS-V3.md)).
 - Ontology **15** or second re-index.
-- Ranking / incremental rebuild proposes (`RANKING-MICROSERVICE`, `TIER2-INCREMENTAL-REBUILD`).
+- Ranking / incremental rebuild proposals.
 - Special-casing `tests/bank-chat-system/` in production code.
 
 # Whole-plan done definition

@@ -96,9 +96,9 @@ Staleness paragraph (step 3, after the `After two failed attempts…` line):
 
 ```markdown
 **Staleness rule:** after `java-codebase-rag increment`, Lance is fresh
-but Kuzu may be stale (see `propose/TIER2-INCREMENTAL-REBUILD-PROPOSE.md`).
-A graph older than the source tree is normal mid-development. When in
-doubt, run `meta` and compare against your working tree.
+but Kuzu may be stale. A graph older than the source tree is normal
+mid-development. When in doubt, run `meta` and compare against your
+working tree.
 ```
 
 **Why the placement**: the recovery playbook is where agents land after a confusing result; the staleness note reads as a follow-on to the existing “stop thrashing” sentence, not as a stray table footer.
@@ -260,9 +260,9 @@ disagreement as evidence of staleness, not as a contradiction.
 
 ```markdown
 **Staleness rule:** after `java-codebase-rag increment`, Lance is fresh
-but Kuzu may be stale (see `propose/TIER2-INCREMENTAL-REBUILD-PROPOSE.md`).
-A graph older than the source tree is normal mid-development. When in
-doubt, run `meta` and compare against your working tree.
+but Kuzu may be stale. A graph older than the source tree is normal
+mid-development. When in doubt, run `meta` and compare against your
+working tree.
 ```
 
 ### A.4 — Patch C (after `**Batching:**` under `#### neighbors`)

@@ -15,7 +15,7 @@
 - **Rename `lancedb_data` default index directory** to `.java-codebase-rag/` (a hidden dotted directory under the project root, matching the dotted-config-file convention). All hardcoded `./lancedb_data` defaults updated.
 - **Rename Python package** `user_rag/` → `java_codebase_rag/`. Update `pyproject.toml [project.scripts]`, all imports, the test file `tests/test_user_rag_cli.py`, and the `python -m user_rag.cli` example in the operator playbook. Set `pyproject.toml [project].name` to **`java-codebase-rag`** (aligned with the GitHub repository basename).
 - One-release deprecation: `refresh` stays as a hidden alias for `reprocess` with a stderr warning, then drops. **No deprecation window for env vars / config / package rename** — breaking changes allowed (no users yet).
-- Migration shape: **3 PRs** — propose merge → CLI + config consolidation (this design) → docs update across the full tree. Engine work for true Kuzu incremental is out of scope and tracked under `propose/TIER2-INCREMENTAL-REBUILD-PROPOSE.md`.
+- Migration shape: **3 PRs** — propose merge → CLI + config consolidation (this design) → docs update across the full tree. Engine work for true Kuzu incremental is out of scope.
 
 ---
 
@@ -230,7 +230,7 @@ Three more renames that the env-var work makes the natural moment for:
 
 ### 3.9 GitHub issue for Kuzu incremental tracking
 
-Open a GitHub issue titled **"AST graph (Kuzu) incremental rebuild"** referencing `propose/TIER2-INCREMENTAL-REBUILD-PROPOSE.md`. Issue URL is hard-coded into the `increment` warning message. The issue is the user-facing handle the warning points at; the propose is the technical scope. Use the **canonical GitHub repo path for this project** when the issue is filed (see Appendix A); do not hard-code a stale org/repo slug in the merged propose.
+Open a GitHub issue titled **"AST graph (Kuzu) incremental rebuild"**. Issue URL is hard-coded into the `increment` warning message. The issue is the user-facing handle the warning points at. Use the **canonical GitHub repo path for this project** when the issue is filed (see Appendix A); do not hard-code a stale org/repo slug in the merged code.
 
 ---
 
@@ -280,7 +280,7 @@ Open a GitHub issue titled **"AST graph (Kuzu) incremental rebuild"** referencin
 
 | Question / feature | Why we skip it |
 |---|---|
-| Kuzu incremental rebuild | Engine work; tracked under `propose/TIER2-INCREMENTAL-REBUILD-PROPOSE.md` and the new tracking issue. Out of scope of a CLI restructure. |
+| Kuzu incremental rebuild | Engine work; tracked under the GitHub issue #73. Out of scope of a CLI restructure. |
 | Auto-detect "should this be `init` or `reprocess`?" | Violates principle 3 (one scenario, one safe default). Operator intent matters; refusing on `init` and being explicit on `reprocess` is the safety policy. |
 | Pipeline-stage subcommands (`build-graph`, `update-lance`) | Violates principle 1. Direct script invocation (`python build_ast_graph.py`) is the escape hatch for advanced users; documented in `docs/JAVA-CODEBASE-RAG-CLI.md` but not a first-class CLI verb. |
 | Graph-freshness query (`meta --graph-freshness`) | Useful but distinct surface; should ride with the Tier-2 incremental work where graph version tracking gets implemented. |
@@ -314,8 +314,6 @@ Open a GitHub issue titled **"AST graph (Kuzu) incremental rebuild"** referencin
 - `.cursor/rules/*.mdc` — agent workflow / env / CLI contract; see **Agent rules audit** below (must match post-rename surface).
 - `docs/CODEBASE_REQUIREMENTS.md` — every `.lancedb-mcp.yml` / `LANCEDB_MCP_*` / `lancedb_data` reference updated.
 - `mcp.json.example` — **PR-CLI-3 is a second pass only:** PR-CLI-2 updates this file so **env keys match the live server**; PR-CLI-3 reconciles comments, examples, and any doc drift — **no conflicting edits**; if both PRs touch it, **PR-CLI-2 wins** for structure, PR-CLI-3 for prose polish.
-- `propose/INDEX-AUTO-MODE-PROPOSE.md` — one-line note that `refresh` is being renamed to `reprocess`.
-- `propose/TIER2-INCREMENTAL-REBUILD-PROPOSE.md` — one-line note that the new tracking issue (created in PR-CLI-2) is the user-facing handle.
 - `docs/PRODUCT-VISION.md` — update `lancedb_data` mention (§ about Kuzu's on-disk footprint) and any `refresh` reference.
 - `.gitignore` — add `.java-codebase-rag/`, keep `lancedb_data/` for grace-period cleanup, or remove if PR-CLI-2 drops that grace-period entry.