fix openenv.yaml with spec_version, type, runtime, app, port

openenv.yaml

@@ -1,83 +1,77 @@
+---
+title: SQL Correction RL Environment
+emoji: 🛢️
+colorFrom: blue
+colorTo: indigo
+sdk: docker
+pinned: false
+tags:
+  - openenv
+---
+
 spec_version: 1
 name: sql-correction-env
-version: "1.0.0"
+type: sequential
+runtime: docker
+app: server.app:app
+port: 7860
+
 description: >
   An OpenEnv RL environment where an AI agent fixes broken SQL queries.
   Simulates a real developer task: identifying and correcting SQL syntax
   and logical errors across easy, medium, and hard difficulty levels.
+
 author: SyncShift
-tags:
-  - openenv
-  - sql
-  - code-correction
-  - real-world
+
+tasks:
+  - name: easy
+    difficulty: easy
+    max_steps: 5
+    description: Fix a single syntax error. Error hint provided.
+    grader: sql_env.grader.grade
+
+  - name: medium
+    difficulty: medium
+    max_steps: 5
+    description: Fix multiple errors across keywords and clauses. No hint.
+    grader: sql_env.grader.grade
+
+  - name: hard
+    difficulty: hard
+    max_steps: 4
+    description: Fix many errors in complex multi-join queries. Schema provided.
+    grader: sql_env.grader.grade
 
 observation_space:
   type: object
   fields:
     task_id:
       type: string
-      description: Unique identifier for the current task
     broken_query:
       type: string
-      description: The malformed SQL query the agent must fix
     schema_context:
       type: string
       nullable: true
-      description: Table and column definitions when a task includes schema context
     error_hint:
       type: string
       nullable: true
-      description: A hint describing the type of error (provided on easy tasks only)
     step_number:
       type: integer
-      description: Current step index within the episode
     previous_attempt:
       type: string
       nullable: true
-      description: The agent's last submitted corrected query
     feedback:
       type: string
       nullable: true
-      description: Grader feedback from the previous step
 
 action_space:
   type: object
   fields:
     corrected_query:
       type: string
-      description: The agent's corrected SQL query
 
 reward:
   range: [0.001, 0.999]
   description: >
-    1.0 = exact match, 0.7 = right tokens minor structure diff,
-    0.4 = most keywords correct, 0.2 = basic structure present, 0.0 = invalid SQL.
-    Stagnation penalty of 0.1 applied after 2 consecutive identical rewards.
-
-tasks:
-  - name: easy
-    difficulty: easy
-    max_steps: 5
-    description: Fix a single syntax error. Error hint provided.
-    grader: sql_env.grader.grade
-
-  - name: medium
-    difficulty: medium
-    max_steps: 5
-    description: Fix multiple errors across keywords and clauses. No hint.
-    grader: sql_env.grader.grade
-
-  - name: hard
-    difficulty: hard
-    max_steps: 4
-    description: Fix many errors in complex multi-join queries. Schema provided. No hint.
-    grader: sql_env.grader.grade
-
-endpoints:
-  reset: POST /reset
-  step:  POST /step
-  state: POST /state
-  tasks: GET /tasks
-  grader: POST /grader
-  health: GET /health
+    0.999 = exact match, 0.7 = right tokens minor structure diff,
+    0.4 = most keywords correct, 0.2 = basic structure present, 0.001 = invalid SQL.