Refactor storage: remove Redis duplication, use proper data tiers

- Recipes: Now content-addressed only (cache + IPFS), removed Redis storage
- Runs: Completed runs stored in PostgreSQL, Redis only for task_id mapping
- Add list_runs_by_actor() to database.py for paginated run queries
- Add list_by_type() to cache_manager for filtering by node_type
- Fix upload endpoint to return size and filename fields
- Fix recipe run endpoint with proper DAG input binding
- Fix get_run_service() dependency to pass database module

Storage architecture:
- Redis: Ephemeral only (sessions, task mappings with TTL)
- PostgreSQL: Permanent records (completed runs, metadata)
- Cache: Content-addressed files (recipes, media, outputs)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

app/routers/cache.py

@@ -219,7 +219,12 @@ async def upload_content(
     if error:
         raise HTTPException(400, error)
 
-    return {"content_hash": content_hash, "uploaded": True}
+    return {
+        "content_hash": content_hash,
+        "filename": file.filename,
+        "size": len(content),
+        "uploaded": True,
+    }
 
 
 # Media listing endpoint

app/routers/recipes.py

@@ -172,24 +172,56 @@ async def run_recipe(
     if not recipe:
         raise HTTPException(404, "Recipe not found")
 
-    # Create run using run service
-    run_service = RunService(database, get_redis_client(), get_cache_manager())
-    run, error = await run_service.create_run(
-        recipe=recipe.get("name", recipe_id),
-        inputs=req.inputs,
-        use_dag=True,
-        actor_id=ctx.actor_id,
-        l2_server=ctx.l2_server,
-    )
+    try:
+        import json
 
-    if error:
-        raise HTTPException(400, error)
+        # Create run using run service
+        run_service = RunService(database, get_redis_client(), get_cache_manager())
 
-    return {
-        "run_id": run.run_id,
-        "status": run.status,
-        "message": "Recipe execution started",
-    }
+        # If recipe has a DAG definition, bind inputs and convert to JSON
+        recipe_dag = recipe.get("dag")
+        dag_json = None
+        if recipe_dag and isinstance(recipe_dag, dict):
+            # Bind inputs to the DAG's source nodes
+            dag_copy = json.loads(json.dumps(recipe_dag))  # Deep copy
+            nodes = dag_copy.get("nodes", {})
+
+            # Map input names to content hashes
+            for input_name, content_hash in req.inputs.items():
+                if input_name in nodes:
+                    node = nodes[input_name]
+                    if node.get("type") == "SOURCE":
+                        if "config" not in node:
+                            node["config"] = {}
+                        node["config"]["content_hash"] = content_hash
+
+            dag_json = json.dumps(dag_copy)
+
+        run, error = await run_service.create_run(
+            recipe=recipe.get("name", recipe_id),
+            inputs=req.inputs,
+            use_dag=True,
+            dag_json=dag_json,
+            actor_id=ctx.actor_id,
+            l2_server=ctx.l2_server,
+        )
+
+        if error:
+            raise HTTPException(400, error)
+
+        if not run:
+            raise HTTPException(500, "Run creation returned no result")
+
+        return {
+            "run_id": run.run_id,
+            "status": run.status,
+            "message": "Recipe execution started",
+        }
+    except HTTPException:
+        raise
+    except Exception as e:
+        logger.exception(f"Error running recipe {recipe_id}")
+        raise HTTPException(500, f"Run failed: {e}")
 
 
 @router.get("/{recipe_id}/dag")