fleet-python 0.2.66b2__py3-none-any.whl → 0.2.105__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (70) hide show
  1. examples/export_tasks.py +16 -5
  2. examples/export_tasks_filtered.py +245 -0
  3. examples/fetch_tasks.py +230 -0
  4. examples/import_tasks.py +140 -8
  5. examples/iterate_verifiers.py +725 -0
  6. fleet/__init__.py +128 -5
  7. fleet/_async/__init__.py +27 -3
  8. fleet/_async/base.py +24 -9
  9. fleet/_async/client.py +938 -41
  10. fleet/_async/env/client.py +60 -3
  11. fleet/_async/instance/client.py +52 -7
  12. fleet/_async/models.py +15 -0
  13. fleet/_async/resources/api.py +200 -0
  14. fleet/_async/resources/sqlite.py +1801 -46
  15. fleet/_async/tasks.py +122 -25
  16. fleet/_async/verifiers/bundler.py +22 -21
  17. fleet/_async/verifiers/verifier.py +25 -19
  18. fleet/agent/__init__.py +32 -0
  19. fleet/agent/gemini_cua/Dockerfile +45 -0
  20. fleet/agent/gemini_cua/__init__.py +10 -0
  21. fleet/agent/gemini_cua/agent.py +759 -0
  22. fleet/agent/gemini_cua/mcp/main.py +108 -0
  23. fleet/agent/gemini_cua/mcp_server/__init__.py +5 -0
  24. fleet/agent/gemini_cua/mcp_server/main.py +105 -0
  25. fleet/agent/gemini_cua/mcp_server/tools.py +178 -0
  26. fleet/agent/gemini_cua/requirements.txt +5 -0
  27. fleet/agent/gemini_cua/start.sh +30 -0
  28. fleet/agent/orchestrator.py +854 -0
  29. fleet/agent/types.py +49 -0
  30. fleet/agent/utils.py +34 -0
  31. fleet/base.py +34 -9
  32. fleet/cli.py +1061 -0
  33. fleet/client.py +1060 -48
  34. fleet/config.py +1 -1
  35. fleet/env/__init__.py +16 -0
  36. fleet/env/client.py +60 -3
  37. fleet/eval/__init__.py +15 -0
  38. fleet/eval/uploader.py +231 -0
  39. fleet/exceptions.py +8 -0
  40. fleet/instance/client.py +53 -8
  41. fleet/instance/models.py +1 -0
  42. fleet/models.py +303 -0
  43. fleet/proxy/__init__.py +25 -0
  44. fleet/proxy/proxy.py +453 -0
  45. fleet/proxy/whitelist.py +244 -0
  46. fleet/resources/api.py +200 -0
  47. fleet/resources/sqlite.py +1845 -46
  48. fleet/tasks.py +113 -20
  49. fleet/utils/__init__.py +7 -0
  50. fleet/utils/http_logging.py +178 -0
  51. fleet/utils/logging.py +13 -0
  52. fleet/utils/playwright.py +440 -0
  53. fleet/verifiers/bundler.py +22 -21
  54. fleet/verifiers/db.py +985 -1
  55. fleet/verifiers/decorator.py +1 -1
  56. fleet/verifiers/verifier.py +25 -19
  57. {fleet_python-0.2.66b2.dist-info → fleet_python-0.2.105.dist-info}/METADATA +28 -1
  58. fleet_python-0.2.105.dist-info/RECORD +115 -0
  59. {fleet_python-0.2.66b2.dist-info → fleet_python-0.2.105.dist-info}/WHEEL +1 -1
  60. fleet_python-0.2.105.dist-info/entry_points.txt +2 -0
  61. tests/test_app_method.py +85 -0
  62. tests/test_expect_exactly.py +4148 -0
  63. tests/test_expect_only.py +2593 -0
  64. tests/test_instance_dispatch.py +607 -0
  65. tests/test_sqlite_resource_dual_mode.py +263 -0
  66. tests/test_sqlite_shared_memory_behavior.py +117 -0
  67. fleet_python-0.2.66b2.dist-info/RECORD +0 -81
  68. tests/test_verifier_security.py +0 -427
  69. {fleet_python-0.2.66b2.dist-info → fleet_python-0.2.105.dist-info}/licenses/LICENSE +0 -0
  70. {fleet_python-0.2.66b2.dist-info → fleet_python-0.2.105.dist-info}/top_level.txt +0 -0
examples/export_tasks.py CHANGED
@@ -17,8 +17,7 @@ def main():
17
17
  parser.add_argument(
18
18
  "--task-keys",
19
19
  "-t",
20
- nargs="+",
21
- help="Optional list of task keys to export (space-separated)",
20
+ help="Optional list of task keys to export (comma-separated)",
22
21
  default=None,
23
22
  )
24
23
  parser.add_argument(
@@ -27,6 +26,12 @@ def main():
27
26
  help="Optional task project key to filter tasks",
28
27
  default=None,
29
28
  )
29
+ parser.add_argument(
30
+ "--env-key",
31
+ "-e",
32
+ help="Optional environment key to filter tasks",
33
+ default=None,
34
+ )
30
35
  parser.add_argument(
31
36
  "--output",
32
37
  "-o",
@@ -42,12 +47,13 @@ def main():
42
47
  args.project_key is not None,
43
48
  args.task_keys is not None,
44
49
  args.task_project_key is not None,
50
+ args.env_key is not None,
45
51
  ]
46
52
  )
47
53
 
48
54
  if filters_specified > 1:
49
55
  parser.error(
50
- "Cannot specify multiple filters. Use only one of --project-key, --task-keys, or --task-project-key."
56
+ "Cannot specify multiple filters. Use only one of --project-key, --task-keys, --task-project-key, or --env-key."
51
57
  )
52
58
 
53
59
  # Get account info
@@ -59,13 +65,18 @@ def main():
59
65
  print(f"Loading tasks from project: {args.project_key}")
60
66
  tasks = fleet.load_tasks(project_key=args.project_key)
61
67
  elif args.task_keys:
68
+ # Split comma-separated task keys and strip whitespace
69
+ task_keys_list = [key.strip() for key in args.task_keys.split(",")]
62
70
  print(
63
- f"Loading {len(args.task_keys)} specific task(s): {', '.join(args.task_keys)}"
71
+ f"Loading {len(task_keys_list)} specific task(s): {', '.join(task_keys_list)}"
64
72
  )
65
- tasks = fleet.load_tasks(keys=args.task_keys)
73
+ tasks = fleet.load_tasks(keys=task_keys_list)
66
74
  elif args.task_project_key:
67
75
  print(f"Loading tasks from task project: {args.task_project_key}")
68
76
  tasks = fleet.load_tasks(task_project_key=args.task_project_key)
77
+ elif args.env_key:
78
+ print(f"Loading tasks from environment: {args.env_key}")
79
+ tasks = fleet.load_tasks(env_key=args.env_key)
69
80
  else:
70
81
  print("Loading all tasks")
71
82
  tasks = fleet.load_tasks()
@@ -0,0 +1,245 @@
1
+ """
2
+ Export tasks to JSON, excluding tasks from targets marked as 'unused'.
3
+
4
+ This script filters out tasks whose task_project_target has status='unused',
5
+ ensuring that broken/invalid targets don't pollute exports.
6
+
7
+ Usage:
8
+ python export_tasks_filtered.py --task-project-key my-project
9
+ python export_tasks_filtered.py --project-key my-project --output tasks.json
10
+ python export_tasks_filtered.py --env-key my-env
11
+ """
12
+
13
+ import argparse
14
+ import json
15
+ import os
16
+ from typing import List, Set
17
+
18
+ import fleet
19
+ from dotenv import load_dotenv
20
+ from supabase import create_client, Client
21
+
22
+ load_dotenv()
23
+
24
+
25
+ def get_unused_target_ids(supabase: Client, team_id: str) -> Set[str]:
26
+ """Fetch all target IDs that have status='unused' for the given team."""
27
+ # Get all task_projects for this team first
28
+ projects_response = (
29
+ supabase.table("task_projects")
30
+ .select("id")
31
+ .eq("team_id", team_id)
32
+ .execute()
33
+ )
34
+
35
+ if not projects_response.data:
36
+ return set()
37
+
38
+ project_ids = [p["id"] for p in projects_response.data]
39
+
40
+ # Get all targets with status='unused' for these projects
41
+ targets_response = (
42
+ supabase.table("task_project_targets")
43
+ .select("id")
44
+ .in_("project_id", project_ids)
45
+ .eq("status", "unused")
46
+ .execute()
47
+ )
48
+
49
+ if not targets_response.data:
50
+ return set()
51
+
52
+ return {t["id"] for t in targets_response.data}
53
+
54
+
55
+ def get_task_target_mapping(supabase: Client, task_keys: List[str], team_id: str) -> dict:
56
+ """Fetch task_project_target_id for each task key."""
57
+ if not task_keys:
58
+ return {}
59
+
60
+ # Batch the queries to avoid hitting limits
61
+ BATCH_SIZE = 100
62
+ mapping = {}
63
+
64
+ for i in range(0, len(task_keys), BATCH_SIZE):
65
+ batch_keys = task_keys[i:i + BATCH_SIZE]
66
+ response = (
67
+ supabase.table("eval_tasks")
68
+ .select("key, task_project_target_id")
69
+ .in_("key", batch_keys)
70
+ .eq("team_id", team_id)
71
+ .execute()
72
+ )
73
+
74
+ for row in response.data or []:
75
+ mapping[row["key"]] = row.get("task_project_target_id")
76
+
77
+ return mapping
78
+
79
+
80
+ def main():
81
+ parser = argparse.ArgumentParser(
82
+ description="Export tasks to JSON, excluding tasks from 'unused' targets"
83
+ )
84
+ parser.add_argument(
85
+ "--project-key",
86
+ "-p",
87
+ help="Optional project key to filter tasks",
88
+ default=None,
89
+ )
90
+ parser.add_argument(
91
+ "--task-keys",
92
+ "-t",
93
+ nargs="+",
94
+ help="Optional list of task keys to export (space-separated)",
95
+ default=None,
96
+ )
97
+ parser.add_argument(
98
+ "--task-project-key",
99
+ "-tpk",
100
+ help="Optional task project key to filter tasks",
101
+ default=None,
102
+ )
103
+ parser.add_argument(
104
+ "--env-key",
105
+ "-e",
106
+ help="Optional environment key to filter tasks",
107
+ default=None,
108
+ )
109
+ parser.add_argument(
110
+ "--output",
111
+ "-o",
112
+ help="Output JSON filename (defaults to {team_id}_filtered.json)",
113
+ default=None,
114
+ )
115
+ parser.add_argument(
116
+ "--include-unused",
117
+ action="store_true",
118
+ help="Include tasks from 'unused' targets (disables filtering)",
119
+ )
120
+
121
+ args = parser.parse_args()
122
+
123
+ # Validate that only one filter is specified
124
+ filters_specified = sum(
125
+ [
126
+ args.project_key is not None,
127
+ args.task_keys is not None,
128
+ args.task_project_key is not None,
129
+ args.env_key is not None,
130
+ ]
131
+ )
132
+
133
+ if filters_specified > 1:
134
+ parser.error(
135
+ "Cannot specify multiple filters. Use only one of --project-key, --task-keys, --task-project-key, or --env-key."
136
+ )
137
+
138
+ # Get account info
139
+ account = fleet.env.account()
140
+ print(f"Exporting from team: {account.team_name}")
141
+
142
+ # Initialize Supabase client for filtering
143
+ supabase_url = os.getenv("SUPABASE_URL") or os.getenv("NEXT_PUBLIC_SUPABASE_URL")
144
+ supabase_key = os.getenv("SUPABASE_SERVICE_ROLE_KEY") or os.getenv("SUPABASE_KEY")
145
+
146
+ if not supabase_url or not supabase_key:
147
+ print("⚠ Warning: SUPABASE_URL/SUPABASE_KEY not set - cannot filter by target status")
148
+ print(" Falling back to unfiltered export")
149
+ supabase = None
150
+ else:
151
+ supabase = create_client(supabase_url, supabase_key)
152
+
153
+ # Load tasks
154
+ if args.project_key:
155
+ print(f"Loading tasks from project: {args.project_key}")
156
+ tasks = fleet.load_tasks(project_key=args.project_key)
157
+ elif args.task_keys:
158
+ print(f"Loading {len(args.task_keys)} specific task(s): {', '.join(args.task_keys)}")
159
+ tasks = fleet.load_tasks(keys=args.task_keys)
160
+ elif args.task_project_key:
161
+ print(f"Loading tasks from task project: {args.task_project_key}")
162
+ tasks = fleet.load_tasks(task_project_key=args.task_project_key)
163
+ elif args.env_key:
164
+ print(f"Loading tasks from environment: {args.env_key}")
165
+ tasks = fleet.load_tasks(env_key=args.env_key)
166
+ else:
167
+ print("Loading all tasks")
168
+ tasks = fleet.load_tasks()
169
+
170
+ print(f"\nFound {len(tasks)} task(s) before filtering")
171
+
172
+ # Filter out tasks from unused targets
173
+ filtered_tasks = tasks
174
+ excluded_count = 0
175
+
176
+ if supabase and not args.include_unused:
177
+ print("\nFiltering out tasks from 'unused' targets...")
178
+
179
+ # Get unused target IDs
180
+ unused_target_ids = get_unused_target_ids(supabase, account.team_id)
181
+
182
+ if unused_target_ids:
183
+ print(f" Found {len(unused_target_ids)} unused target(s)")
184
+
185
+ # Get task -> target mapping
186
+ task_keys = [t.key for t in tasks]
187
+ task_target_map = get_task_target_mapping(supabase, task_keys, account.team_id)
188
+
189
+ # Filter tasks
190
+ filtered_tasks = []
191
+ for task in tasks:
192
+ target_id = task_target_map.get(task.key)
193
+ if target_id in unused_target_ids:
194
+ excluded_count += 1
195
+ else:
196
+ filtered_tasks.append(task)
197
+
198
+ print(f" Excluded {excluded_count} task(s) from unused targets")
199
+ else:
200
+ print(" No unused targets found - all tasks included")
201
+
202
+ tasks = filtered_tasks
203
+ print(f"\n{len(tasks)} task(s) after filtering")
204
+
205
+ # Validate that all tasks have verifier_func
206
+ print("\nValidating tasks have verifier_func...")
207
+ missing_verifier = []
208
+ for task in tasks:
209
+ if not task.verifier_func:
210
+ missing_verifier.append(task.key)
211
+
212
+ if missing_verifier:
213
+ print(f"\n✗ Error: {len(missing_verifier)} task(s) missing verifier_func:")
214
+ for key in missing_verifier[:10]: # Show first 10
215
+ print(f" - {key}")
216
+ if len(missing_verifier) > 10:
217
+ print(f" ... and {len(missing_verifier) - 10} more")
218
+ raise ValueError(
219
+ "All tasks must have a verifier_func. Cannot export tasks without verifiers."
220
+ )
221
+
222
+ print("✓ All tasks have verifier_func")
223
+
224
+ # Determine output filename
225
+ output_file = args.output or f"{account.team_id}_filtered.json"
226
+
227
+ # Export to JSON
228
+ print(f"\nExporting to: {output_file}")
229
+ with open(output_file, "w", encoding="utf-8") as f:
230
+ json.dump(
231
+ [task.model_dump() for task in tasks],
232
+ f,
233
+ indent=2,
234
+ ensure_ascii=False,
235
+ )
236
+
237
+ print(f"✓ Successfully exported {len(tasks)} task(s) to {output_file}")
238
+ if excluded_count > 0:
239
+ print(f" ({excluded_count} task(s) excluded from unused targets)")
240
+
241
+
242
+ if __name__ == "__main__":
243
+ main()
244
+
245
+
@@ -0,0 +1,230 @@
1
+ import argparse
2
+ import asyncio
3
+ import json
4
+ import sys
5
+ from typing import List, Dict, Any, Optional, Tuple
6
+ import fleet
7
+ from dotenv import load_dotenv
8
+
9
+ load_dotenv()
10
+
11
+
12
+ async def fetch_task(
13
+ task_key: str, semaphore: asyncio.Semaphore
14
+ ) -> Tuple[str, Optional[Dict[str, Any]], Optional[str]]:
15
+ """
16
+ Fetch a single task from the Fleet API.
17
+
18
+ Args:
19
+ task_key: Task key to fetch
20
+ semaphore: Semaphore to limit concurrent requests
21
+
22
+ Returns:
23
+ Tuple of (task_key, task_data_dict, error_message)
24
+ """
25
+ async with semaphore:
26
+ try:
27
+ # Use load_tasks with keys parameter to get Task objects
28
+ tasks = await fleet.load_tasks_async(keys=[task_key])
29
+ if tasks:
30
+ task = tasks[0]
31
+ # Convert to dict using model_dump() like export_tasks.py does
32
+ return task_key, task.model_dump(), None
33
+ else:
34
+ return task_key, None, "Task not found"
35
+ except Exception as e:
36
+ error_msg = f"{type(e).__name__}: {str(e)}"
37
+ return task_key, None, error_msg
38
+
39
+
40
+ async def fetch_tasks_batch(
41
+ task_keys: List[str], max_concurrent: int = 20
42
+ ) -> Dict[str, Dict[str, Any]]:
43
+ """
44
+ Fetch multiple tasks concurrently from the Fleet API.
45
+
46
+ Args:
47
+ task_keys: List of task keys to fetch
48
+ max_concurrent: Maximum number of concurrent requests
49
+
50
+ Returns:
51
+ Dictionary mapping task_key to task data
52
+ """
53
+ print(f"\nFetching {len(task_keys)} task(s) from Fleet API...")
54
+ print(f"Max concurrent requests: {max_concurrent}")
55
+
56
+ semaphore = asyncio.Semaphore(max_concurrent)
57
+
58
+ # Fetch all tasks concurrently
59
+ results = await asyncio.gather(
60
+ *[fetch_task(key, semaphore) for key in task_keys],
61
+ return_exceptions=True,
62
+ )
63
+
64
+ # Process results
65
+ fetched_tasks = {}
66
+ errors = []
67
+
68
+ for result in results:
69
+ if isinstance(result, Exception):
70
+ errors.append(f"Unexpected error: {result}")
71
+ continue
72
+
73
+ task_key, task_data, error = result
74
+
75
+ if error:
76
+ errors.append(f"{task_key}: {error}")
77
+ print(f" ✗ {task_key}: {error}")
78
+ elif task_data:
79
+ # Task data is already a dict from model_dump()
80
+ fetched_tasks[task_key] = task_data
81
+ print(f" ✓ {task_key}")
82
+
83
+ print(f"\n✓ Successfully fetched {len(fetched_tasks)} task(s)")
84
+
85
+ if errors:
86
+ print(f"\n⚠ {len(errors)} error(s) occurred:")
87
+ for error in errors[:10]: # Show first 10
88
+ print(f" - {error}")
89
+ if len(errors) > 10:
90
+ print(f" ... and {len(errors) - 10} more")
91
+
92
+ return fetched_tasks
93
+
94
+
95
+ async def main():
96
+ parser = argparse.ArgumentParser(
97
+ description="Fetch tasks from Fleet API and update JSON file",
98
+ formatter_class=argparse.RawDescriptionHelpFormatter,
99
+ epilog="""
100
+ Examples:
101
+ # Fetch tasks and update in-place
102
+ %(prog)s tasks.json
103
+
104
+ # Fetch tasks and save to a new file
105
+ %(prog)s tasks.json --output updated_tasks.json
106
+
107
+ # Limit concurrent requests
108
+ %(prog)s tasks.json --max-concurrent 10
109
+ """,
110
+ )
111
+
112
+ parser.add_argument("json_file", help="Path to JSON file containing tasks")
113
+ parser.add_argument(
114
+ "--output",
115
+ "-o",
116
+ help="Path to output JSON file (defaults to overwriting input)",
117
+ default=None,
118
+ )
119
+ parser.add_argument(
120
+ "--max-concurrent",
121
+ "-c",
122
+ type=int,
123
+ default=20,
124
+ help="Maximum number of concurrent API requests (default: 20)",
125
+ )
126
+
127
+ args = parser.parse_args()
128
+
129
+ # Load JSON file
130
+ print(f"Reading tasks from: {args.json_file}")
131
+ try:
132
+ with open(args.json_file, "r", encoding="utf-8") as f:
133
+ tasks_data = json.load(f)
134
+ except FileNotFoundError:
135
+ print(f"✗ Error: File '{args.json_file}' not found")
136
+ sys.exit(1)
137
+ except json.JSONDecodeError as e:
138
+ print(f"✗ Error: Invalid JSON in '{args.json_file}': {e}")
139
+ sys.exit(1)
140
+
141
+ if not isinstance(tasks_data, list):
142
+ print("✗ Error: JSON file must contain an array of tasks")
143
+ sys.exit(1)
144
+
145
+ print(f"Found {len(tasks_data)} task(s) in file")
146
+
147
+ # Extract task keys
148
+ task_keys = []
149
+ missing_keys = []
150
+
151
+ for i, task in enumerate(tasks_data):
152
+ task_key = task.get("key") or task.get("id")
153
+ if task_key:
154
+ task_keys.append(task_key)
155
+ else:
156
+ missing_keys.append(f"Task at index {i}")
157
+
158
+ if missing_keys:
159
+ print(f"\n⚠ Warning: {len(missing_keys)} task(s) missing key/id:")
160
+ for key in missing_keys[:10]:
161
+ print(f" - {key}")
162
+ if len(missing_keys) > 10:
163
+ print(f" ... and {len(missing_keys) - 10} more")
164
+
165
+ if not task_keys:
166
+ print("\n✗ Error: No valid task keys found in JSON file")
167
+ sys.exit(1)
168
+
169
+ print(f"\nExtracted {len(task_keys)} task key(s)")
170
+
171
+ # Get account info
172
+ account = await fleet.env.account_async()
173
+ print(f"Fetching from team: {account.team_name}")
174
+
175
+ # Fetch tasks from API
176
+ fetched_tasks = await fetch_tasks_batch(task_keys, args.max_concurrent)
177
+
178
+ if not fetched_tasks:
179
+ print("\n✗ Error: No tasks were successfully fetched")
180
+ sys.exit(1)
181
+
182
+ # Update tasks in the original data
183
+ updated_count = 0
184
+ not_found = []
185
+
186
+ print("\nUpdating task data...")
187
+ for i, task in enumerate(tasks_data):
188
+ task_key = task.get("key") or task.get("id")
189
+
190
+ if task_key in fetched_tasks:
191
+ # Replace entire task with fetched data
192
+ tasks_data[i] = fetched_tasks[task_key]
193
+ updated_count += 1
194
+ else:
195
+ not_found.append(task_key or f"index {i}")
196
+
197
+ print(f"✓ Updated {updated_count} task(s)")
198
+
199
+ if not_found:
200
+ print(f"\n⚠ Warning: {len(not_found)} task(s) not fetched:")
201
+ for key in not_found[:10]:
202
+ print(f" - {key}")
203
+ if len(not_found) > 10:
204
+ print(f" ... and {len(not_found) - 10} more")
205
+
206
+ # Write output
207
+ output_file = args.output or args.json_file
208
+ print(f"\nWriting updated tasks to: {output_file}")
209
+
210
+ try:
211
+ with open(output_file, "w", encoding="utf-8") as f:
212
+ json.dump(tasks_data, f, indent=2, ensure_ascii=False)
213
+ print(f"✓ Successfully wrote {len(tasks_data)} task(s) to '{output_file}'")
214
+ except Exception as e:
215
+ print(f"✗ Error writing output file: {e}")
216
+ sys.exit(1)
217
+
218
+ # Summary
219
+ print("\n" + "=" * 60)
220
+ print("Summary:")
221
+ print(f" Total tasks in file: {len(tasks_data)}")
222
+ print(f" Successfully fetched: {len(fetched_tasks)}")
223
+ print(f" Updated in file: {updated_count}")
224
+ print(f" Failed to fetch: {len(task_keys) - len(fetched_tasks)}")
225
+ print("=" * 60)
226
+
227
+
228
+ if __name__ == "__main__":
229
+ asyncio.run(main())
230
+