## 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.
141 lines
6.0 KiB
Python
141 lines
6.0 KiB
Python
"""Every AgentLoop construction site must hand over the capabilities it owns.
|
|
|
|
``TokenWiseConfig.enabled``, ``.cache_optimization`` and ``.max_cache_breakpoints``
|
|
all defaulted to working values, ``CacheOptimizer`` was complete and benchmarked,
|
|
and ``install_from_config`` -- the one function that turns those flags into a
|
|
registry -- had no callers. Nothing passed ``strategies=`` to ``AgentLoop``, so
|
|
every surface ran the empty pass-through registry and the strategy never
|
|
executed once.
|
|
|
|
The default value was correct, so a test over defaults, a schema assertion or a
|
|
config validator cannot catch this: it was a reachability failure, not a
|
|
configuration one. The only thing that catches the class is enumerating the
|
|
construction sites and requiring each to either wire the capability or say in
|
|
writing why it does not.
|
|
|
|
That is what this file does. It is deliberately source-level, because the
|
|
failure mode is the absence of an argument -- which produces no runtime signal
|
|
anywhere.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import ast
|
|
from pathlib import Path
|
|
|
|
import pytest
|
|
|
|
RAVEN = Path(__file__).resolve().parents[1] / "raven"
|
|
|
|
# A construction site may skip a capability only with a reason recorded here.
|
|
# Adding an entry is a deliberate act; forgetting an argument is not.
|
|
_EXEMPT: dict[tuple[str, str], str] = {
|
|
("trajectory/replay.py", "strategies"): (
|
|
"Replay re-runs a recorded trajectory to compare against what was recorded. "
|
|
"Placing live cache breakpoints would change the request being replayed, so "
|
|
"the comparison would no longer be against the run it is reproducing."
|
|
),
|
|
}
|
|
|
|
# Each kwarg here carries an operator setting whose absence has no runtime
|
|
# symptom, because the loop's own default is a working value.
|
|
_REQUIRED = ("strategies",)
|
|
|
|
|
|
def _flat_kwargs(call: ast.Call) -> set[str]:
|
|
"""Keyword names a site passes, descending into the wiring bundles."""
|
|
flat: set[str] = set()
|
|
for kw in call.keywords:
|
|
inner = kw.value
|
|
if isinstance(inner, ast.Call) and getattr(inner.func, "id", "").endswith(("Wiring", "Policy")):
|
|
flat |= {k.arg for k in inner.keywords if k.arg}
|
|
elif kw.arg:
|
|
flat.add(kw.arg)
|
|
return flat
|
|
|
|
|
|
def _sites() -> list[tuple[str, ast.Call]]:
|
|
found: list[tuple[str, ast.Call]] = []
|
|
for path in sorted(RAVEN.rglob("*.py")):
|
|
if "__pycache__" in path.parts:
|
|
continue
|
|
tree = ast.parse(path.read_text(encoding="utf-8"))
|
|
for node in ast.walk(tree):
|
|
if not isinstance(node, ast.Call):
|
|
continue
|
|
func = node.func
|
|
name = func.id if isinstance(func, ast.Name) else getattr(func, "attr", None)
|
|
if name == "AgentLoop":
|
|
found.append((path.relative_to(RAVEN).as_posix(), node))
|
|
return found
|
|
|
|
|
|
def test_the_enumeration_still_finds_the_construction_sites():
|
|
"""The negative assertions below all pass over an empty list."""
|
|
names = {rel for rel, _ in _sites()}
|
|
assert "core/runtime.py" in names, f"the assembly door stopped building an AgentLoop by name: {sorted(names)}"
|
|
|
|
|
|
@pytest.mark.parametrize("capability", _REQUIRED)
|
|
def test_every_construction_site_wires_every_capability(capability: str):
|
|
missing = [
|
|
rel for rel, call in _sites() if capability not in _flat_kwargs(call) and (rel, capability) not in _EXEMPT
|
|
]
|
|
assert not missing, (
|
|
f"these AgentLoop sites do not pass {capability!r}: {missing}\n"
|
|
f"Wire it, or add an entry to _EXEMPT with the reason."
|
|
)
|
|
|
|
|
|
def test_the_probe_each_site_passes_reads_the_running_turns_binding():
|
|
"""A probe closed over the construction-time provider is wrong per session.
|
|
|
|
``AgentLoop.provider`` is a property over the active binding, so a session
|
|
that switched models runs on a provider the loop was not built with. The
|
|
strategies run inside ``use_binding``, and the answer has to follow them
|
|
there -- otherwise a session switched onto a caching model is told it cannot
|
|
cache, and one switched off it marks a request that cannot carry the field.
|
|
"""
|
|
from raven.core.token_wise_stack import caching_probe
|
|
from raven.providers.binding import ModelBinding, use_binding
|
|
from raven.providers.litellm_provider import LiteLLMProvider
|
|
|
|
model = "openrouter/anthropic/claude-fable-5"
|
|
built_with = LiteLLMProvider(api_key="k", default_model=model, provider_name="custom")
|
|
switched_to = LiteLLMProvider(api_key="k", default_model=model, provider_name="openrouter")
|
|
probe = caching_probe(built_with)
|
|
|
|
assert probe(model) is False, "outside a turn, the default binding is the only answer"
|
|
with use_binding(ModelBinding(provider=switched_to, model=model)):
|
|
assert probe(model) is True, "inside a turn, the session's own wire decides"
|
|
|
|
|
|
def test_no_production_caller_redirects_usage_telemetry():
|
|
"""``settings.usage`` scans exactly one directory: UsageTracker's default.
|
|
|
|
A caller passing ``telemetry_dir=`` writes rows that reader never sees --
|
|
shipped once as ``workspace/.token_wise``, where default-on usage tracking
|
|
reported ``calls=0`` in settings. Like the ``strategies=`` check above,
|
|
the defect is an argument with no runtime symptom, so the only thing that
|
|
catches the class is enumerating the call sites.
|
|
"""
|
|
offenders = [
|
|
rel for rel, call in _calls("install_from_config") if "telemetry_dir" in {kw.arg for kw in call.keywords}
|
|
]
|
|
assert not offenders, (
|
|
f"these install_from_config sites redirect usage telemetry away from "
|
|
f"the directory settings.usage reads: {offenders}"
|
|
)
|
|
|
|
|
|
def _calls(func_name: str) -> list[tuple[str, ast.Call]]:
|
|
found: list[tuple[str, ast.Call]] = []
|
|
for path in sorted(RAVEN.rglob("*.py")):
|
|
if "__pycache__" in path.parts:
|
|
continue
|
|
tree = ast.parse(path.read_text(encoding="utf-8"))
|
|
for node in ast.walk(tree):
|
|
if isinstance(node, ast.Call) and isinstance(node.func, ast.Name) and node.func.id == func_name:
|
|
found.append((path.relative_to(RAVEN).as_posix(), node))
|
|
return found
|