## Summary
Reported as "deliver_files is broken". It is not: **the tool never ran.**
Every `deliver_files` call on one install was refused before dispatch -- five in a row in a single session -- and nothing was ever delivered:
```
Error: Invalid parameters for tool 'deliver_files': files should be array
```
What the model sent was the array, spelled as a string:
```json
{"files": "[{\"path\": \"...poster.png\", \"title\": \"...\"}]", "message": "..."}
```
`cast_params` already forgives that spelling for every scalar type -- `"3"` for an integer, `"true"` for a boolean -- and did not for arrays or objects, which are exactly the types a model is most likely to stringify.
## Type
- [x] Fix
- [ ] Feature
## The edge of the forgiving is the design
When the value is a string, try `json.loads`, and take the result **only if it is the type the schema asked for**. A string that is not JSON, or is JSON of the wrong shape, is returned unchanged so the validator still names the real problem -- the point is to widen what is accepted, not to swallow a type error.
**Nothing here can raise.** Turning "your argument has the wrong type" into a traceback would be worse than the refusal it replaces, so the parse is wrapped and failure falls through.
**A `type: "string"` field is never parsed.** JSON is ordinary text in a message body or a file's content, and replacing it would hand the tool something it never asked for. Two independent things prevent it today (the string branch returns before the parse is reached, and the parse only wins on a type match), which is why the case pinning it is a tripwire rather than a load-bearing assertion -- said so in its own docstring rather than left to look like coverage.
## Not just deliver_files
Six built-in tools declare an array or object parameter and were all exposed:
| tool | parameter |
|---|---|
| `deliver_files` | `files` |
| `ask_user` | `questions` |
| `image_generate` | `images` |
| `message` | `media` |
| `create_playbook` | `skills` |
| `video_generate` | `params` |
The same failure was already seen from the other side: `acp_client/acp_agent.py` records `update_task_state` and `operations should be array` from a real dispatch.
## Verification
```
make lint-python ruff check + ruff format --check, 1850 files
make lint-imports 9 contracts kept, 0 broken
pytest tests/ 19,699 passed, 7 failed (all pre-existing, see below)
```
- [x] Relevant tests pass locally
- [x] Relevant lint / type checks pass locally
- [ ] User-facing docs or screenshots -- no user-facing string
**End to end, through the real tool**, with the exact payload that failed five times in production:
```
files sent as: str -> [{"path": ".../poster.png", "title": "..."}]
returned: Delivered 1 file(s): poster.png (72 B).
still "should be array": False
registered in the store: True
```
Twelve cases, five mutations, each killing at least one:
| mutation | red |
|---|---|
| drop the whole branch (the old behaviour) | 3 |
| raise instead of falling through on a parse failure | 5 |
| accept the parse whatever type it yielded | 1 |
| do not re-cast the parsed value, so items are not cast | 1 |
| widen the tuple to include `string` | **survived -- unreachable** |
The last is worth stating rather than hiding: that mutation cannot change behaviour, because the `type == "string"` branch returns earlier and the type check would refuse the result anyway. I also tried moving the new block above that branch, alone and combined with widening the tuple, and neither reddens anything either. The invariant holds by construction; the case that documents it says so.
**Pre-existing failures, checked against the base.** 7 failed in the full run, all in `tests/test_security_network.py` (`test_a_redirect_to_a_dangerous_address_is_refused` and siblings). Running that file alone on this same tree gives 4 -- the count is order-dependent -- and the change touches only `raven/agent/tools/params.py` plus a new test file, neither of which `raven/security/network.py` imports. The failures are the box's resolver: it returns poisoned answers roughly half the time, which I measured separately (`en.wikipedia.org` came back as `2001::1` on 3 of 6 lookups). Not introduced here, and worth someone's attention on its own.
## Risk
- [x] Security impact considered -- widens what reaches a tool's own validation, not what bypasses it. A parsed value is still validated against the same schema, and a parse that yields the wrong type is refused exactly as before
- [x] Backward compatibility considered -- a call that worked before takes the identical path; the new branch is reached only by a `str` where an array or object is declared, which previously always failed
- [x] Rollback path is clear -- single-commit revert
## Related Issues
N/A
## Swept
`mr-review-patterns` pre-submit sweep, over `git diff origin/main...HEAD`.
- **1 (a write location moved)** -- nothing moved; a branch was added to an existing dispatch on `target_type`. Grepped every caller of `cast_params`: the registry is the only one, and it runs `validate_params` right after, which is the reader this affects.
- **2 (a value with several entry points)** -- the cast is the single funnel every tool argument passes through, which is why the fix belongs here and not in `deliver_files`.
- **3 (a guard)** -- checked both directions: what it wrongly lets through (JSON of the wrong type, non-JSON text -- both refused, with cases) and what it wrongly refuses (a real list, unaffected, with a case).
- **4 (tests)** -- five mutations above, one honestly reported as unreachable. Failing-ID sets diffed against the base rather than counts.
- **6 (stale text)** -- the module docstring enumerated what is coerced ("a string that should be an integer"); it now names arrays and objects too. Grepped the tracked tree for other statements of the rule: none.
- **9 (error handling)** -- the new path adds one way to end and it is caught: `json.loads` raises `ValueError` on bad text and `TypeError` on a non-string, and both fall through to the untouched value.
- **11** -- `git ls-files | grep -c node_modules` is 0; the commit-message gate was run, not eyeballed.
Not applicable, with reasons: **3b, 3c, 3d, 3e** (no tool schema produced, no state moved, one lane, no assertion lost its symbol); **4b, 4c, 4d** (no platform branch, no shell, no rendered number); **5, 5b** (no dependency; `json` is stdlib and already imported across this package); **6b, 6c** (no backup, no generated artifact); **7, 7b** (nothing on disk); **8** (no asset, no secret); **10** (no prompt text).
## Deliberately not fixed
The error text still ends with `[Analyze the error above and try a different approach.]`, and the model plainly did not -- it sent the same shape five times. A message naming the expected spelling (`files must be a JSON array, not a string containing one`) would give it something to act on. That is a change to the registry's error surface, affects every tool, and deserves its own argument.
## Moved from !594
Opened on a compliant source branch. AGENTS.md 2.1 asks for a 3-5 word snake_case description and the previous branch had seven; GitLab cannot change an existing merge request's source branch, so the change moves rather than the name.
Same commits, rebased onto `c2a9768f`. Both code findings from the review on !594 are already in them -- `RecursionError` caught with its regression case, and the implementation comment reduced to the two non-obvious reasons. Nothing else changed, so that review carries.
446 lines
14 KiB
Python
446 lines
14 KiB
Python
from __future__ import annotations
|
|
|
|
import json
|
|
|
|
import pytest
|
|
|
|
from raven.agent import workdir
|
|
from raven_design.task_state import TaskStateError, TaskStateManager
|
|
from raven_design.task_state.tool import TaskStateTool
|
|
|
|
|
|
def _state() -> dict:
|
|
return {
|
|
"goal": "Write a report",
|
|
"requirements": ["Keep it concise"],
|
|
"items": [
|
|
{
|
|
"title": "Collect results",
|
|
"requirements": ["Read all result files"],
|
|
},
|
|
{
|
|
"title": "Write report",
|
|
"requirements": ["Save the final report"],
|
|
},
|
|
],
|
|
}
|
|
|
|
|
|
def test_initialize_renders_visible_numbers_without_ids(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
|
|
rendered = manager.render("cli:one")
|
|
|
|
assert "1. [pending] Collect results" in rendered
|
|
assert "2. [pending] Write report" in rendered
|
|
assert "item_number" in rendered
|
|
assert "item_id" not in rendered
|
|
|
|
|
|
def test_item_number_updates_the_snapshot_seen_by_the_model(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
|
|
manager.apply(
|
|
"cli:one",
|
|
[{"operation": "update", "item_number": 2, "changes": {"status": "in_progress"}}],
|
|
)
|
|
|
|
assert manager.get("cli:one")["items"][1]["status"] == "in_progress"
|
|
|
|
|
|
def test_batch_operations_resolve_numbers_against_the_original_snapshot(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
|
|
manager.apply(
|
|
"cli:one",
|
|
[
|
|
{"operation": "remove", "item_number": 1, "reason": "Merged into the report step."},
|
|
{"operation": "complete", "item_number": 2},
|
|
],
|
|
)
|
|
|
|
state = manager.get("cli:one")
|
|
assert len(state["items"]) == 1
|
|
assert state["items"][0]["title"] == "Write report"
|
|
assert state["items"][0]["status"] == "completed"
|
|
|
|
|
|
def test_failed_batch_does_not_write_partial_changes(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
before = manager.revision("cli:one")
|
|
|
|
with pytest.raises(TaskStateError):
|
|
manager.apply(
|
|
"cli:one",
|
|
[
|
|
{"operation": "update", "item_number": 1, "changes": {"status": "in_progress"}},
|
|
{"operation": "complete", "item_number": 99},
|
|
],
|
|
)
|
|
|
|
assert manager.revision("cli:one") == before
|
|
assert manager.get("cli:one")["items"][0]["status"] == "pending"
|
|
|
|
|
|
def test_corrupt_storage_degrades_and_quarantines(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
path = manager.store._path("cli:one")
|
|
path.write_text("{ not json", encoding="utf-8")
|
|
|
|
assert manager.get("cli:one") is None
|
|
assert manager.revision("cli:one") == 0
|
|
assert manager.render("cli:one").startswith("<task_state>\nRevision: 0\nNot initialized")
|
|
assert not path.exists()
|
|
assert path.with_name(path.name + ".corrupt").exists()
|
|
|
|
update = manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
assert update.revision == 1
|
|
|
|
|
|
def test_structurally_invalid_state_degrades_instead_of_raising(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
path = manager.store._path("cli:one")
|
|
path.write_text(
|
|
json.dumps({"version": 1, "revision": 3, "state": {"goal": "x"}}),
|
|
encoding="utf-8",
|
|
)
|
|
|
|
assert manager.get("cli:one") is None
|
|
assert manager.snapshot("cli:one") is None
|
|
assert path.with_name(path.name + ".corrupt").exists()
|
|
|
|
|
|
def test_operation_rejects_unknown_fields_without_writing(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
before = manager.revision("cli:one")
|
|
|
|
with pytest.raises(TaskStateError, match="Unknown operation fields"):
|
|
manager.apply(
|
|
"cli:one",
|
|
[{"operation": "complete", "item_number": 1, "unexpected": "value"}],
|
|
)
|
|
|
|
assert manager.revision("cli:one") == before
|
|
assert manager.get("cli:one")["items"][0]["status"] == "pending"
|
|
|
|
|
|
def test_initialize_accepts_placeholders_for_other_operation_shapes(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
|
|
manager.apply(
|
|
"cli:one",
|
|
[
|
|
{
|
|
"operation": "initialize",
|
|
"state": _state(),
|
|
"changes": {
|
|
"goal": "",
|
|
"requirements": [],
|
|
"status": "pending",
|
|
"title": "",
|
|
},
|
|
"item": {
|
|
"requirements": [],
|
|
"status": "pending",
|
|
"title": "",
|
|
},
|
|
"item_number": 1,
|
|
"reason": "",
|
|
"target": "task_state",
|
|
}
|
|
],
|
|
)
|
|
|
|
assert manager.get("cli:one")["goal"] == "Write a report"
|
|
|
|
|
|
def test_update_ignores_explicit_null_placeholders(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
|
|
manager.apply(
|
|
"cli:one",
|
|
[
|
|
{
|
|
"operation": "update",
|
|
"target": "task_state",
|
|
"changes": {"goal": "Updated goal", "requirements": None},
|
|
},
|
|
{
|
|
"operation": "update",
|
|
"target": "item",
|
|
"item_number": 1,
|
|
"changes": {
|
|
"title": "Updated item",
|
|
"status": "in_progress",
|
|
"requirements": None,
|
|
},
|
|
},
|
|
],
|
|
)
|
|
|
|
state = manager.get("cli:one")
|
|
assert state["goal"] == "Updated goal"
|
|
assert state["requirements"] == ["Keep it concise"]
|
|
assert state["items"][0] == {
|
|
"title": "Updated item",
|
|
"status": "in_progress",
|
|
"requirements": ["Read all result files"],
|
|
}
|
|
|
|
|
|
def test_tool_schema_separates_operation_shapes(tmp_path):
|
|
schema = TaskStateTool(TaskStateManager(tmp_path)).parameters
|
|
branches = schema["properties"]["operations"]["anyOf"]
|
|
assert branches[0]["maxItems"] == 1
|
|
variants = [branches[0]["items"], *branches[1]["items"]["oneOf"]]
|
|
|
|
assert [variant["properties"]["operation"]["enum"][0] for variant in variants] == [
|
|
"initialize",
|
|
"add",
|
|
"update",
|
|
"update",
|
|
"remove",
|
|
"complete",
|
|
]
|
|
assert set(variants[2]["properties"]) == {"operation", "target", "changes"}
|
|
assert set(variants[3]["properties"]) == {
|
|
"operation",
|
|
"target",
|
|
"item_number",
|
|
"changes",
|
|
}
|
|
|
|
|
|
def test_initialize_replaces_the_whole_state(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
replacement = {
|
|
"goal": "New goal",
|
|
"requirements": [],
|
|
"items": [{"title": "Only step", "status": "pending", "requirements": []}],
|
|
}
|
|
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": replacement}])
|
|
|
|
assert manager.get("cli:one") == replacement
|
|
|
|
|
|
def test_initialize_cannot_be_mixed_with_incremental_operations(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
|
|
with pytest.raises(TaskStateError, match="must be called alone"):
|
|
manager.apply(
|
|
"cli:one",
|
|
[
|
|
{"operation": "initialize", "state": _state()},
|
|
{"operation": "add", "item": {"title": "Extra"}},
|
|
],
|
|
)
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_tool_uses_the_visible_item_number(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
tool = TaskStateTool(manager)
|
|
with workdir.bind(tmp_path):
|
|
await tool.execute([{"operation": "initialize", "state": _state()}])
|
|
|
|
result = await tool.execute(
|
|
[{"operation": "complete", "item_number": 1}],
|
|
)
|
|
|
|
assert "1/2 completed" in result
|
|
assert manager.get(str(tmp_path))["items"][0]["status"] == "completed"
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_tool_rejects_unknown_top_level_fields(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
tool = TaskStateTool(manager)
|
|
with workdir.bind(tmp_path):
|
|
result = await tool.execute(
|
|
[{"operation": "initialize", "state": _state()}],
|
|
unexpected="value",
|
|
)
|
|
|
|
assert result == "Error: Unknown Task State fields: ['unexpected']"
|
|
assert manager.get(str(tmp_path)) is None
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_undo_reverts_all_mutations_from_one_turn(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
manager.apply(
|
|
"cli:one",
|
|
[{"operation": "update", "item_number": 1, "changes": {"status": "in_progress"}}],
|
|
mutation_group="turn-one",
|
|
)
|
|
manager.apply(
|
|
"cli:one",
|
|
[{"operation": "update", "item_number": 2, "changes": {"status": "in_progress"}}],
|
|
mutation_group="turn-one",
|
|
)
|
|
manager.undo("cli:one")
|
|
|
|
assert [item["status"] for item in manager.get("cli:one")["items"]] == ["pending", "pending"]
|
|
|
|
|
|
def test_undo_tracks_a_turn_that_did_not_mutate_task_state(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
|
|
manager.begin_turn("cli:one", "turn-one")
|
|
manager.apply(
|
|
"cli:one",
|
|
[{"operation": "update", "item_number": 1, "changes": {"status": "in_progress"}}],
|
|
mutation_group="turn-one",
|
|
)
|
|
manager.begin_turn("cli:one", "turn-two")
|
|
|
|
manager.undo("cli:one")
|
|
|
|
assert manager.get("cli:one")["items"][0]["status"] == "in_progress"
|
|
|
|
|
|
def test_undo_can_restore_an_uninitialized_state(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
|
|
manager.begin_turn("cli:one", "turn-one")
|
|
manager.apply(
|
|
"cli:one",
|
|
[{"operation": "initialize", "state": _state()}],
|
|
mutation_group="turn-one",
|
|
)
|
|
|
|
manager.undo("cli:one")
|
|
|
|
assert manager.get("cli:one") is None
|
|
|
|
|
|
def test_initialize_without_a_turn_group_is_also_undoable(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
|
|
manager.undo("cli:one")
|
|
|
|
assert manager.get("cli:one") is None
|
|
|
|
|
|
def test_copy_preserves_rollback_history(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
manager.apply(
|
|
"cli:one",
|
|
[{"operation": "update", "item_number": 1, "changes": {"status": "in_progress"}}],
|
|
)
|
|
|
|
manager.copy("cli:one", "cli:branch")
|
|
manager.undo("cli:branch")
|
|
|
|
assert manager.get("cli:branch")["items"][0]["status"] == "pending"
|
|
assert manager.get("cli:one")["items"][0]["status"] == "in_progress"
|
|
|
|
|
|
def test_state_file_is_json_and_persistent(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
|
|
payload = json.loads(next((tmp_path / "task_states").glob("*.json")).read_text())
|
|
|
|
assert payload["revision"] == 1
|
|
assert payload["state"]["goal"] == "Write a report"
|
|
|
|
|
|
def test_unicode_session_key_uses_a_portable_bounded_filename(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
session_key = f"channel:{'任务🚀' * 100}"
|
|
|
|
manager.apply(session_key, [{"operation": "initialize", "state": _state()}])
|
|
|
|
path = next((tmp_path / "task_states").glob("*.json"))
|
|
assert len(path.name.encode()) <= 255
|
|
assert manager.get(session_key)["goal"] == "Write a report"
|
|
|
|
|
|
def test_initialize_can_replace_with_no_items(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply(
|
|
"cli:one",
|
|
[{"operation": "initialize", "state": {"goal": "Answer one question", "requirements": [], "items": []}}],
|
|
)
|
|
|
|
assert manager.get("cli:one")["items"] == []
|
|
assert manager.unfinished_numbers("cli:one") == []
|
|
|
|
|
|
def test_undo_restores_the_previous_state(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
manager.apply(
|
|
"cli:one",
|
|
[{"operation": "update", "item_number": 1, "changes": {"status": "in_progress"}}],
|
|
)
|
|
|
|
manager.undo("cli:one")
|
|
|
|
assert manager.get("cli:one")["items"][0]["status"] == "pending"
|
|
|
|
|
|
def test_snapshot_contains_only_visible_item_numbers(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
manager.apply("cli:one", [{"operation": "initialize", "state": _state()}])
|
|
|
|
snapshot = manager.snapshot("cli:one")
|
|
|
|
assert snapshot["items"][0]["item_number"] == 1
|
|
assert "item_id" not in snapshot["items"][0]
|
|
|
|
|
|
@pytest.mark.parametrize("extra", ["add", "initialize"])
|
|
def test_schema_rejects_initialization_batched_with_any_operation(tmp_path, extra):
|
|
from jsonschema import Draft202012Validator
|
|
|
|
schema = TaskStateTool(TaskStateManager(tmp_path)).parameters
|
|
initialize = {"operation": "initialize", "state": _state()}
|
|
other = {"operation": "add", "item": {"title": "Extra"}} if extra == "add" else initialize
|
|
validator = Draft202012Validator(schema)
|
|
assert validator.is_valid({"operations": [initialize]})
|
|
assert not validator.is_valid({"operations": [initialize, other]})
|
|
assert not validator.is_valid({"operations": [other, initialize]})
|
|
assert validator.is_valid(
|
|
{
|
|
"operations": [
|
|
{"operation": "add", "item": {"title": "Extra"}},
|
|
{"operation": "complete", "item_number": 1},
|
|
]
|
|
}
|
|
)
|
|
|
|
|
|
def test_mixed_initialization_error_explains_how_to_retry(tmp_path):
|
|
manager = TaskStateManager(tmp_path)
|
|
with pytest.raises(TaskStateError, match="all initial tasks in state.items"):
|
|
manager.apply(
|
|
"test",
|
|
[
|
|
{"operation": "initialize", "state": _state()},
|
|
{"operation": "add", "item": {"title": "Extra"}},
|
|
],
|
|
)
|
|
assert manager.get("test") is None
|
|
manager.apply("test", [{"operation": "initialize", "state": _state()}])
|
|
manager.apply(
|
|
"test", [{"operation": "add", "item": {"title": "Extra"}}, {"operation": "complete", "item_number": 1}]
|
|
)
|
|
assert len(manager.get("test")["items"]) == 3
|
|
assert manager.get("test")["items"][0]["status"] == "completed"
|