Store Cognee datasets in the projects bucket

CHANGELOG.md

@@ -10,7 +10,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### ✨ Enhancements
 - Added Ladybug-backed Cognee integration with optional MinIO/S3 storage. Projects can now set `COGNEE_STORAGE_BACKEND=s3` (plus the `COGNEE_S3_*` settings) to keep knowledge graphs in the shared MinIO bucket seeded by `docker-compose`, enabling multi-tenant ingestion across workers and containers.
 - Introduced a dedicated Cognee service (`docker/docker-compose.cognee.yml`) and HTTP client so `fuzzforge ingest` streams data to the shared backend (`COGNEE_SERVICE_URL`) instead of importing Cognee locally. Each project now auto-provisions its own Cognee account/tenant and authenticates via the REST API, keeping datasets isolated even though the service is shared.
-- Added an event-driven ingestion pipeline: MinIO publishes `PUT` events from `s3://cognee/projects/<project-id>/...` to RabbitMQ, and the new `ingestion-dispatcher` container downloads the file, logs into Cognee as that project’s tenant, and invokes `/api/v1/add` + `/api/v1/cognify`. Uploading files (rsync, CI, etc.) now keeps datasets fresh without touching the CLI.
+- Added an event-driven ingestion pipeline: MinIO publishes `PUT` events from `s3://projects/<project-id>/...` to RabbitMQ, and the new `ingestion-dispatcher` container downloads the file, logs into Cognee as that project’s tenant, and invokes `/api/v1/add` + `/api/v1/cognify`. Uploading files (rsync, CI, etc.) now keeps datasets fresh without touching the CLI.
 
 ### 📝 Documentation
 - Added comprehensive worker startup documentation across all guides

README.md

@@ -197,7 +197,7 @@ ff workflow run security_assessment .    # Start workflow - CLI uploads files au
 Uploading files into MinIO automatically streams them into Cognee:
 
 ```
-s3://cognee/projects/<project-id>/
+s3://projects/<project-id>/
   files/...        # → <project-id>_codebase dataset
   findings/...     # → <project-id>_findings dataset
   docs/...         # → <project-id>_docs dataset

cli/src/fuzzforge_cli/config.py

@@ -138,7 +138,7 @@ class CogneeConfig(BaseModel):
     service_password: Optional[str] = None
     storage_backend: Literal["local", "s3"] = "s3"
     s3_bucket: Optional[str] = None
-    s3_prefix: Optional[str] = "projects"
+    s3_prefix: Optional[str] = None
     s3_endpoint_url: Optional[str] = None
     s3_region: Optional[str] = None
     s3_access_key: Optional[str] = None
@@ -217,8 +217,12 @@ class FuzzForgeConfig(BaseModel):
             cognee.service_url = "http://localhost:18000"
             changed = True
 
-        if not cognee.s3_prefix:
-            cognee.s3_prefix = "projects"
+        if not cognee.s3_bucket:
+            cognee.s3_bucket = "projects"
+            changed = True
+
+        if cognee.s3_prefix is None:
+            cognee.s3_prefix = ""
             changed = True
 
         default_email = f"project_{self.project.id}@fuzzforge.dev"
@@ -234,9 +238,13 @@ class FuzzForgeConfig(BaseModel):
             changed = True
 
         if cognee.storage_backend.lower() == "s3":
-            bucket = cognee.s3_bucket or "cognee"
-            prefix = (cognee.s3_prefix or "projects").strip("/")
-            base_uri = f"s3://{bucket}/{prefix}/{self.project.id}"
+            bucket = cognee.s3_bucket or "projects"
+            prefix = (cognee.s3_prefix or "").strip("/")
+            path_parts = [f"s3://{bucket}"]
+            if prefix:
+                path_parts.append(prefix)
+            path_parts.append(self.project.id)
+            base_uri = "/".join(path_parts)
             data_dir = f"{base_uri}/files"
             system_dir = f"{base_uri}/graph"
         else:

docker/docker-compose.cognee.yml

@@ -12,10 +12,7 @@ services:
       GRAPH_DATABASE_PROVIDER: ladybug
       ENABLE_BACKEND_ACCESS_CONTROL: "true"
       STORAGE_BACKEND: s3
-      STORAGE_BUCKET_NAME: ${COGNEE_S3_BUCKET:-cognee}
-      DATA_ROOT_DIRECTORY: s3://${COGNEE_S3_BUCKET:-cognee}/${COGNEE_S3_PREFIX:-projects}
-      SYSTEM_ROOT_DIRECTORY: s3://${COGNEE_S3_BUCKET:-cognee}/${COGNEE_S3_PREFIX:-projects}
-      CACHE_ROOT_DIRECTORY: s3://${COGNEE_S3_BUCKET:-cognee}/${COGNEE_S3_PREFIX:-projects}/cache
+      STORAGE_BUCKET_NAME: ${COGNEE_S3_BUCKET:-projects}
       DB_PROVIDER: sqlite
       DB_PATH: /data/relational
       DB_NAME: cognee.db

docs/docs/ai/a2a-services.md

@@ -12,7 +12,7 @@ Run the command from a project directory that already contains `.fuzzforge/`. Th
 
 **Default directories / services**
 - Logs: `.fuzzforge/logs/cognee.log`
-- Cognee datasets: hosted by the shared Cognee service (`COGNEE_SERVICE_URL`) inside the configured MinIO/S3 bucket. Local mode falls back to `.fuzzforge/cognee/project_<id>/{data,system}`. Uploads dropped into `s3://cognee/projects/<project-id>/...` are ingested automatically via RabbitMQ + the dispatcher.
+- Cognee datasets: hosted by the shared Cognee service (`COGNEE_SERVICE_URL`) inside the configured MinIO/S3 bucket. Local mode falls back to `.fuzzforge/cognee/project_<id>/{data,system}`. Uploads dropped into `s3://projects/<project-id>/...` are ingested automatically via RabbitMQ + the dispatcher.
 - Artifact cache: `.fuzzforge/artifacts`
 
 ## HTTP Endpoints

docs/docs/ai/architecture.md

@@ -92,7 +92,7 @@ The CLI surface mirrors these helpers as natural-language prompts (`You> run fuz
 
 ## Knowledge & Ingestion
 
-- The `fuzzforge ingest` and `fuzzforge rag ingest` commands call into `ai/src/fuzzforge_ai/ingest_utils.py`, which filters file types, ignores caches, and streams files to the Cognee service where they are stored under `s3://<bucket>/<prefix>/project_<id>/`. When files land directly in `s3://cognee/projects/<project-id>/<category>/...`, the dispatcher performs the same workflow automatically via RabbitMQ events.
+- The `fuzzforge ingest` and `fuzzforge rag ingest` commands call into `ai/src/fuzzforge_ai/ingest_utils.py`, which filters file types, ignores caches, and streams files to the Cognee service where they are stored under `s3://projects/<project-id>/`. When files land directly in `s3://projects/<project-id>/<category>/...`, the dispatcher performs the same workflow automatically via RabbitMQ events.
 - Runtime queries hit `query_project_knowledge_api` on the executor, which defers to `cognee_service` for dataset lookup and semantic search. When Cognee credentials are absent the tools return a friendly "not configured" response.
 
 ## Artifact Pipeline

docs/docs/ai/ingestion.md

@@ -102,7 +102,7 @@ Add comments or project-specific overrides as needed; the agent reads these vari
 
 ## Event-Driven Ingestion
 
-Uploading files directly into MinIO triggers Cognee automatically. The dispatcher watches `s3://cognee/projects/<project-id>/...` and translates the top-level folder into a dataset:
+Uploading files directly into MinIO triggers Cognee automatically. The dispatcher watches `s3://projects/<project-id>/...` and translates the top-level folder into a dataset:
 
 | Prefix     | Dataset name                          |
 |-----------|---------------------------------------|

docs/docs/how-to/docker-setup.md

@@ -243,7 +243,7 @@ This spins up the Cognee API on `http://localhost:18000`, publishes it to the ho
 
 ### RabbitMQ + Dispatcher
 
-`docker-compose.yml` also launches RabbitMQ (`http://localhost:15672`, ingest/ingest) and the `ingestion-dispatcher` container. MinIO publishes `PUT` events from `s3://cognee/projects/<project-id>/...` to the `cognee-ingest` exchange, and the dispatcher downloads the object and calls Cognee’s REST API. That means any rsync/upload into the projects bucket automatically becomes a dataset.
+`docker-compose.yml` also launches RabbitMQ (`http://localhost:15672`, ingest/ingest) and the `ingestion-dispatcher` container. MinIO publishes `PUT` events from `s3://projects/<project-id>/...` to the `cognee-ingest` exchange, and the dispatcher downloads the object and calls Cognee’s REST API. That means any rsync/upload into the projects bucket automatically becomes a dataset.
 
 ---
 

volumes/env/.env.template

@@ -69,8 +69,8 @@ COGNEE_MCP_URL=
 COGNEE_SERVICE_URL=http://localhost:18000
 COGNEE_API_KEY=
 COGNEE_STORAGE_BACKEND=s3
-COGNEE_S3_BUCKET=cognee
-COGNEE_S3_PREFIX=projects
+COGNEE_S3_BUCKET=projects
+COGNEE_S3_PREFIX=
 COGNEE_S3_ENDPOINT=http://localhost:9000
 COGNEE_S3_REGION=us-east-1
 COGNEE_S3_ACCESS_KEY=fuzzforge