rename project from citelens to sourcery

Author: sushildalavi

.github/workflows/ci.yml

@@ -170,10 +170,10 @@ jobs:
           version: v3.16.3
 
       - name: helm lint
-        run: helm lint helm/citelens
+        run: helm lint helm/sourcery
 
       - name: helm template (dry render)
-        run: helm template citelens helm/citelens --debug | head -200
+        run: helm template sourcery helm/sourcery --debug | head -200
 
   pyright:
     name: pyright (typecheck)

CHANGELOG.md

@@ -33,7 +33,7 @@ roughly tracks [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
 ### Fixed
 - Frontend fast-refresh warning by splitting `useDefaultPaletteActions` into its own file.
-- README pointed at the wrong GitHub repo slug (`Final-Project-ScholarRAG` → `citelens`).
+- README pointed at the wrong GitHub repo slug (`Final-Project-ScholarRAG` → `sourcery`).
 
 ## [0.1.0] — initial commit history
 

Makefile

@@ -38,23 +38,23 @@ test-coverage:
 # on port 5433 (so it doesn't fight whatever's already on 5432). Idempotent —
 # safe to re-run; the container is recreated each time.
 test-isolation:
-	@echo "→ tearing down any previous citelens-iso-db"
-	@docker rm -f citelens-iso-db 2>/dev/null || true
+	@echo "→ tearing down any previous sourcery-iso-db"
+	@docker rm -f sourcery-iso-db 2>/dev/null || true
 	@echo "→ booting pgvector on :5433"
-	@docker run -d --name citelens-iso-db \
+	@docker run -d --name sourcery-iso-db \
 		-e POSTGRES_USER=scholarrag -e POSTGRES_PASSWORD=scholarrag \
 		-e POSTGRES_DB=scholarrag -p 5433:5432 \
 		pgvector/pgvector:pg16 >/dev/null
 	@sleep 6
-	@docker exec -i citelens-iso-db psql -U scholarrag -d scholarrag < db/init.sql >/dev/null
+	@docker exec -i sourcery-iso-db psql -U scholarrag -d scholarrag < db/init.sql >/dev/null
 	@echo "→ running isolation tests"
 	@DATABASE_URL=postgresql://scholarrag:scholarrag@127.0.0.1:5433/scholarrag \
 		PGHOST=127.0.0.1 PGPORT=5433 PGUSER=scholarrag PGPASSWORD=scholarrag PGDATABASE=scholarrag \
 		EMBEDDING_PROVIDER=stub OPENAI_API_KEY=test \
 		$(PYTEST) backend/tests/test_workspace_isolation.py -v --tb=short ; \
 		EXITCODE=$$? ; \
 		echo "→ tearing down" ; \
-		docker rm -f citelens-iso-db >/dev/null 2>&1 || true ; \
+		docker rm -f sourcery-iso-db >/dev/null 2>&1 || true ; \
 		exit $$EXITCODE
 
 # ── Lint ──────────────────────────────────────────────────────────────────────

README.md

@@ -1,12 +1,12 @@
-# citelens
+# sourcery
 
-<img src="frontend/favicon.svg" width="56" alt="citelens" align="left" hspace="12" />
+<img src="frontend/favicon.svg" width="56" alt="sourcery" align="left" hspace="12" />
 
 A scholarly RAG app I built because I kept getting plausible-but-uncited answers from off-the-shelf chatbots. It does hybrid retrieval over your uploaded PDFs and six live scholarly APIs, then scores every generated claim for whether it's actually supported by the cited evidence.
 
 <br clear="left"/>
 
-[![CI](https://img.shields.io/github/actions/workflow/status/sushildalavi/citelens/ci.yml?branch=main&label=CI&logo=github)](https://github.com/sushildalavi/citelens/actions/workflows/ci.yml) [![Tests](https://img.shields.io/badge/tests-198%20%E2%80%A2%2092%25%20unit-15803d?logo=pytest&logoColor=white)](https://github.com/sushildalavi/citelens/actions) [![Coverage](https://img.shields.io/badge/coverage-50%25-15803d)](.github/workflows/ci.yml) [![Python](https://img.shields.io/badge/python-3.11-1d4ed8?logo=python&logoColor=white)](https://www.python.org/) [![FastAPI](https://img.shields.io/badge/FastAPI-0.135+-15803d?logo=fastapi&logoColor=white)](https://fastapi.tiangolo.com/) [![React](https://img.shields.io/badge/React-18-1d4ed8?logo=react&logoColor=white)](https://react.dev/) [![pgvector](https://img.shields.io/badge/Postgres%2016-pgvector-336791?logo=postgresql&logoColor=white)](https://github.com/pgvector/pgvector) [![License](https://img.shields.io/badge/License-MIT-15803d)](LICENSE)
+[![CI](https://img.shields.io/github/actions/workflow/status/sushildalavi/sourcery/ci.yml?branch=main&label=CI&logo=github)](https://github.com/sushildalavi/sourcery/actions/workflows/ci.yml) [![Tests](https://img.shields.io/badge/tests-198%20%E2%80%A2%2092%25%20unit-15803d?logo=pytest&logoColor=white)](https://github.com/sushildalavi/sourcery/actions) [![Coverage](https://img.shields.io/badge/coverage-50%25-15803d)](.github/workflows/ci.yml) [![Python](https://img.shields.io/badge/python-3.11-1d4ed8?logo=python&logoColor=white)](https://www.python.org/) [![FastAPI](https://img.shields.io/badge/FastAPI-0.135+-15803d?logo=fastapi&logoColor=white)](https://fastapi.tiangolo.com/) [![React](https://img.shields.io/badge/React-18-1d4ed8?logo=react&logoColor=white)](https://react.dev/) [![pgvector](https://img.shields.io/badge/Postgres%2016-pgvector-336791?logo=postgresql&logoColor=white)](https://github.com/pgvector/pgvector) [![License](https://img.shields.io/badge/License-MIT-15803d)](LICENSE)
 
 [Quick start](#quick-start) · [Architecture](#architecture) · [Benchmarks](#benchmark-results) · [API examples](docs/examples/curl.md) · [Helm chart](helm/README.md) · [Security](SECURITY.md)
 
@@ -499,8 +499,8 @@ dominates the tail; embed/retrieve/rerank stay sub-450 ms even at p99.
 ### 1. Clone and configure
 
 ```bash
-git clone https://github.com/sushildalavi/citelens.git
-cd citelens
+git clone https://github.com/sushildalavi/sourcery.git
+cd sourcery
 cp .env.example .env
 # fill in OPENAI_API_KEY, DATABASE_URL, OLLAMA_BASE_URL
 ```
@@ -546,7 +546,7 @@ make test
 ## Project Structure
 
 ```
-citelens/
+sourcery/
 ├── backend/
 │   ├── app.py                   # FastAPI app — CORS, routers, startup
 │   ├── pdf_ingest.py            # PDF extraction, chunking, pgvector upsert
@@ -780,7 +780,7 @@ Shipped:
 - [x] **Multi-tenant isolation** — `WorkspaceMiddleware` reads `X-Workspace-Id`, sanitises it, pins it on `request.state`. Every INSERT/SELECT/UPDATE/DELETE on `documents`, `chunks`, `chat_sessions`, `user_memory`, `digests`, and `confidence_calibration` filters by `workspace_id`. `db/init.sql` is canonical (fresh deploys are wired); existing deploys upgrade via `_ensure_workspace_columns()` on FastAPI lifespan startup. Cross-tenant isolation is covered by 7 integration tests in `test_workspace_isolation.py` (run via `make test-isolation`).
 - [x] **Batched embedding upserts** — verified the existing path: cache lookup → batched OpenAI / parallel Ollama → bulk `execute_values` insert into both `chunks` and `chunk_embeddings`. A 50-chunk PDF is one OpenAI call + one cache insert + one upsert per table.
 - [x] **Prometheus exposition** — `GET /metrics/prom` returns text/plain in standard Prometheus format. Helm chart includes an optional `ServiceMonitor` if you run Prometheus Operator.
-- [x] **Helm chart** — `helm/citelens/` with backend + frontend Deployments, Postgres StatefulSet, optional Ingress + ServiceMonitor. Wired to the k8s-style `/health/live` + `/health/ready` probes.
+- [x] **Helm chart** — `helm/sourcery/` with backend + frontend Deployments, Postgres StatefulSet, optional Ingress + ServiceMonitor. Wired to the k8s-style `/health/live` + `/health/ready` probes.
 - [x] **Per-tenant calibration** — `_load_latest_calibration_weights(workspace_id=...)` looks up tenant-specific MSA weights in the calibration table and falls back gracefully (tenant→global→default).
 
 Still open:

backend/.env.example

@@ -14,7 +14,7 @@ PG_POOL_MAX=10
 # --- OpenAI (generation + LLM judge) ---
 OPENAI_API_KEY=sk-...
 # If using AWS Secrets Manager instead of plain env:
-# OPENAI_API_KEY_SECRET_ID=prod/citelens/openai
+# OPENAI_API_KEY_SECRET_ID=prod/sourcery/openai
 
 # --- Embeddings ---
 # Provider: ollama (default, local) or openai

backend/app.py

@@ -370,17 +370,17 @@ def _emit(name: str, value, help_text: str, mtype: str = "gauge") -> None:
         lines.append(f"# TYPE {name} {mtype}")
         lines.append(f"{name} {num}")
 
-    _emit("citelens_documents_total", payload.get("documents"), "Documents in `ready` status.")
-    _emit("citelens_chunks_total", payload.get("chunks"), "Total chunks across all documents.")
-    _emit("citelens_eval_runs_total", payload.get("eval_runs"), "Recorded evaluation runs.")
+    _emit("sourcery_documents_total", payload.get("documents"), "Documents in `ready` status.")
+    _emit("sourcery_chunks_total", payload.get("chunks"), "Total chunks across all documents.")
+    _emit("sourcery_eval_runs_total", payload.get("eval_runs"), "Recorded evaluation runs.")
     retrieval = payload.get("retrieval") or {}
-    _emit("citelens_retrieval_recall_at_5", retrieval.get("recall_at_5"), "Recall@5 from latest eval run.")
-    _emit("citelens_retrieval_ndcg_at_10", retrieval.get("ndcg_at_10"), "nDCG@10 from latest eval run.")
-    _emit("citelens_retrieval_mrr", retrieval.get("mrr"), "Mean reciprocal rank from latest eval run.")
+    _emit("sourcery_retrieval_recall_at_5", retrieval.get("recall_at_5"), "Recall@5 from latest eval run.")
+    _emit("sourcery_retrieval_ndcg_at_10", retrieval.get("ndcg_at_10"), "nDCG@10 from latest eval run.")
+    _emit("sourcery_retrieval_mrr", retrieval.get("mrr"), "Mean reciprocal rank from latest eval run.")
     latency = payload.get("latency_ms") or {}
-    _emit("citelens_assistant_latency_ms_p50", latency.get("p50"), "Assistant answer latency p50 (ms).")
-    _emit("citelens_assistant_latency_ms_p95", latency.get("p95"), "Assistant answer latency p95 (ms).")
-    _emit("citelens_assistant_latency_ms_p99", latency.get("p99"), "Assistant answer latency p99 (ms).")
+    _emit("sourcery_assistant_latency_ms_p50", latency.get("p50"), "Assistant answer latency p50 (ms).")
+    _emit("sourcery_assistant_latency_ms_p95", latency.get("p95"), "Assistant answer latency p95 (ms).")
+    _emit("sourcery_assistant_latency_ms_p99", latency.get("p99"), "Assistant answer latency p99 (ms).")
 
     body = "\n".join(lines) + "\n" if lines else ""
     return Response(content=body, media_type="text/plain; version=0.0.4; charset=utf-8")

backend/tests/test_metrics_prometheus.py

@@ -28,12 +28,12 @@ def _fake_metrics():
     body = resp.text
 
     # Spot-check the wire format.
-    assert "# HELP citelens_documents_total" in body
-    assert "# TYPE citelens_documents_total gauge" in body
-    assert "citelens_documents_total 42.0" in body
-    assert "citelens_chunks_total 1337.0" in body
-    assert "citelens_retrieval_recall_at_5 0.99" in body
-    assert "citelens_assistant_latency_ms_p99 1600.0" in body
+    assert "# HELP sourcery_documents_total" in body
+    assert "# TYPE sourcery_documents_total gauge" in body
+    assert "sourcery_documents_total 42.0" in body
+    assert "sourcery_chunks_total 1337.0" in body
+    assert "sourcery_retrieval_recall_at_5 0.99" in body
+    assert "sourcery_assistant_latency_ms_p99 1600.0" in body
 
 
 def test_metrics_prom_skips_missing_values(monkeypatch):
@@ -49,5 +49,5 @@ def test_metrics_prom_skips_missing_values(monkeypatch):
     client = TestClient(app_module.app)
     body = client.get("/metrics/prom").text
 
-    assert "citelens_documents_total" not in body
-    assert "citelens_chunks_total 5.0" in body
+    assert "sourcery_documents_total" not in body
+    assert "sourcery_chunks_total 5.0" in body

backend/tests/test_workspace_isolation.py

@@ -11,11 +11,11 @@
 When DATABASE_URL isn't set they skip with a clear message rather than
 failing. To run by hand:
 
-    docker run -d --name citelens-iso-db \\
+    docker run -d --name sourcery-iso-db \\
       -e POSTGRES_USER=scholarrag -e POSTGRES_PASSWORD=scholarrag \\
       -e POSTGRES_DB=scholarrag -p 5433:5432 pgvector/pgvector:pg16
     sleep 5
-    docker exec -i citelens-iso-db psql -U scholarrag -d scholarrag < db/init.sql
+    docker exec -i sourcery-iso-db psql -U scholarrag -d scholarrag < db/init.sql
     DATABASE_URL=postgresql://scholarrag:scholarrag@127.0.0.1:5433/scholarrag \\
       EMBEDDING_PROVIDER=stub OPENAI_API_KEY=test \\
       pytest backend/tests/test_workspace_isolation.py -v

docs/architecture.md

@@ -1,4 +1,4 @@
-# Architecture — citelens
+# Architecture — sourcery
 
 This document describes the **why** behind the major design decisions. The
 **what** lives in the [README architecture diagrams](../README.md#architecture).

frontend/index.html

@@ -3,10 +3,10 @@
   <head>
     <meta charset="UTF-8" />
     <meta name="viewport" content="width=device-width, initial-scale=1.0" />
-    <meta name="description" content="Citelens — calibrated RAG over your PDFs and 6 live scholarly APIs, with per-claim faithfulness scoring." />
+    <meta name="description" content="Sourcery — calibrated RAG over your PDFs and 6 live scholarly APIs, with per-claim faithfulness scoring." />
     <meta name="theme-color" content="#0f172a" />
     <link rel="icon" type="image/svg+xml" href="/favicon.svg" />
-    <title>Citelens — Calibrated Scholarly RAG</title>
+    <title>Sourcery — Calibrated Scholarly RAG</title>
     <link rel="preconnect" href="https://fonts.googleapis.com" />
     <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin />
     <link