## 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.
215 lines
8.4 KiB
Python
215 lines
8.4 KiB
Python
"""A turn the runtime opened is marked as such in the transcript.
|
|
|
|
Reopening a session drew the announce that ends a sub-agent run as the user's
|
|
own words -- untrusted fence, instance handle, and the instruction saying not to
|
|
repeat either of them to the user, all on screen. The text is what the model
|
|
reads; a reader needs to know a person did not type it.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import json
|
|
|
|
import pytest
|
|
|
|
from raven.agent.loop._shared import _ORIGIN_KEY
|
|
from raven.rpc.methods.session import _map_to_wire
|
|
from raven.spine.turn import Origin
|
|
|
|
|
|
def _stored(**over):
|
|
entry = {"role": "user", "content": "hi", "timestamp": "2026-08-21T15:00:00"}
|
|
entry.update(over)
|
|
return entry
|
|
|
|
|
|
# ── the wire ──────────────────────────────────────────────────────
|
|
|
|
|
|
def test_a_stored_origin_reaches_the_wire() -> None:
|
|
"""`session.resume` copies a whitelist onto the wire entry; a field absent
|
|
from it is dropped however carefully it was written."""
|
|
[entry] = _map_to_wire([_stored(origin="subagent")], "tui:default")
|
|
assert entry["origin"] == "subagent"
|
|
|
|
|
|
def test_an_unmarked_entry_carries_no_origin() -> None:
|
|
[entry] = _map_to_wire([_stored()], "tui:default")
|
|
assert "origin" not in entry
|
|
|
|
|
|
@pytest.mark.parametrize("origin", ["subagent", "cron", "sentinel", "heartbeat"])
|
|
def test_every_runtime_origin_survives_the_wire(origin) -> None:
|
|
"""Not only sub-agents: a cron reminder's text carries "when you reply,
|
|
mention when the reminder was originally set", which is as much an
|
|
instruction to the model as an announce is."""
|
|
[entry] = _map_to_wire([_stored(origin=origin)], "tui:default")
|
|
assert entry["origin"] == origin
|
|
|
|
|
|
def test_the_text_still_rides() -> None:
|
|
"""The model reads it. Only the reader is meant to treat it differently."""
|
|
[entry] = _map_to_wire([_stored(origin="subagent", content="[Subagent 'x' completed]")], "tui:default")
|
|
assert entry["text"] == "[Subagent 'x' completed]"
|
|
|
|
|
|
# ── the storage rename ────────────────────────────────────────────
|
|
|
|
|
|
def test_the_private_key_is_renamed_for_storage(tmp_path) -> None:
|
|
"""Same underscore-then-rename convention as ``_notice``: the underscore
|
|
keeps it out of the live provider payload, storage gets the plain name."""
|
|
from raven.agent.loop.main import AgentLoop
|
|
|
|
class _Session:
|
|
def __init__(self):
|
|
self.entries = []
|
|
|
|
def record(self, entry):
|
|
self.entries.append(entry)
|
|
|
|
loop = object.__new__(AgentLoop)
|
|
loop._now_fn = lambda: __import__("datetime").datetime(2026, 8, 21, 15, 0) # _save_turn stamps entries
|
|
session = _Session()
|
|
messages = [{"role": "user", "content": "hi", _ORIGIN_KEY: "subagent"}]
|
|
AgentLoop._save_turn(loop, session, messages, 0)
|
|
|
|
[stored] = session.entries
|
|
assert stored["origin"] == "subagent"
|
|
assert _ORIGIN_KEY not in stored
|
|
|
|
|
|
def test_an_unmarked_turn_stores_no_origin() -> None:
|
|
from raven.agent.loop.main import AgentLoop
|
|
|
|
class _Session:
|
|
def __init__(self):
|
|
self.entries = []
|
|
|
|
def record(self, entry):
|
|
self.entries.append(entry)
|
|
|
|
loop = object.__new__(AgentLoop)
|
|
loop._now_fn = lambda: __import__("datetime").datetime(2026, 8, 21, 15, 0)
|
|
session = _Session()
|
|
AgentLoop._save_turn(loop, session, [{"role": "user", "content": "hi"}], 0)
|
|
assert "origin" not in session.entries[0]
|
|
|
|
|
|
# ── the contract ──────────────────────────────────────────────────
|
|
|
|
|
|
def test_the_schema_declares_the_field() -> None:
|
|
"""The page's client is generated from this file, so a field the schema does
|
|
not name cannot be read by the reader that needs it."""
|
|
from pathlib import Path
|
|
|
|
schema = json.loads((Path(__file__).resolve().parent.parent / "rpc-schema" / "openrpc.json").read_text())
|
|
props = schema["components"]["schemas"]["TranscriptMessage"]["properties"]
|
|
assert props["origin"]["type"] == "string"
|
|
|
|
|
|
def test_the_origins_named_in_the_description_are_the_real_ones() -> None:
|
|
"""The description lists them for a reader of the contract; a rename in the
|
|
enum would otherwise leave the contract quietly wrong."""
|
|
from pathlib import Path
|
|
|
|
schema = json.loads((Path(__file__).resolve().parent.parent / "rpc-schema" / "openrpc.json").read_text())
|
|
described = schema["components"]["schemas"]["TranscriptMessage"]["properties"]["origin"]["description"]
|
|
for origin in (Origin.SUBAGENT, Origin.CRON, Origin.SENTINEL, Origin.HEARTBEAT):
|
|
assert f"`{origin}`" in described
|
|
|
|
|
|
# ── which turns are marked ────────────────────────────────────────
|
|
|
|
|
|
@pytest.mark.parametrize("origin", [Origin.SUBAGENT, Origin.CRON, Origin.SENTINEL, Origin.HEARTBEAT])
|
|
def test_every_origin_but_the_user_is_marked(origin) -> None:
|
|
"""Marking only sub-agents would leave cron, sentinel and heartbeat drawing
|
|
their runtime prose as typed words -- the same bug, three more ways in."""
|
|
from raven.agent.loop._shared import _runtime_origin
|
|
|
|
assert _runtime_origin(origin) == str(origin)
|
|
|
|
|
|
def test_a_person_typing_is_not_marked() -> None:
|
|
from raven.agent.loop._shared import _runtime_origin
|
|
|
|
assert _runtime_origin(Origin.USER) is None
|
|
assert _runtime_origin(None) is None
|
|
|
|
|
|
# ── the path that answers before _save_turn runs ──────────────────
|
|
# With personalization on, a runtime-origin turn can be met with a clarifying
|
|
# question and returned, so the envelope the assembler would have marked is
|
|
# never built and the entry reaches disk from `_record_inbound` instead.
|
|
|
|
|
|
class _Recorder:
|
|
def __init__(self) -> None:
|
|
self.entries: list[dict] = []
|
|
|
|
def record(self, entry: dict) -> None:
|
|
self.entries.append(entry)
|
|
|
|
|
|
@pytest.mark.parametrize("origin", [Origin.SENTINEL, Origin.CRON, Origin.HEARTBEAT])
|
|
def test_the_clarification_path_marks_what_it_stores(origin) -> None:
|
|
from raven.agent.loop.main import AgentLoop
|
|
|
|
session = _Recorder()
|
|
AgentLoop._record_inbound(object.__new__(AgentLoop), session, "runtime notice", origin, "2026-08-22T15:00:00")
|
|
|
|
[stored] = session.entries
|
|
assert stored["role"] == "user"
|
|
assert stored["origin"] == str(origin)
|
|
|
|
|
|
def test_the_clarification_path_leaves_a_persons_message_unmarked() -> None:
|
|
from raven.agent.loop.main import AgentLoop
|
|
|
|
session = _Recorder()
|
|
AgentLoop._record_inbound(object.__new__(AgentLoop), session, "hi", Origin.USER, "2026-08-22T15:00:00")
|
|
|
|
assert "origin" not in session.entries[0]
|
|
|
|
|
|
def test_no_path_writes_the_inbound_entry_around_the_writer() -> None:
|
|
"""Two copies of this write is how one came to be marked and the other not,
|
|
so the invariant is that the raw form is gone rather than that some function
|
|
contains a call."""
|
|
import inspect
|
|
|
|
from raven.agent.loop import main as loop_main
|
|
from raven.agent.loop import turn_path
|
|
|
|
source = inspect.getsource(turn_path)
|
|
# Exactly one: the writer's own line. A second is a path that skipped it.
|
|
assert source.count('{"role": "user", "content": content') == 1
|
|
assert '{"role": "user", "content": content' in inspect.getsource(loop_main.AgentLoop._record_inbound)
|
|
assert source.count("self._record_inbound(session, content, origin") == 2
|
|
|
|
|
|
# ── the other readers of the same contract ────────────────────────
|
|
|
|
|
|
def test_the_markdown_export_does_not_sign_it_as_the_user() -> None:
|
|
"""`session.export` is read by a person too. Under "User" it asserts that
|
|
someone typed an announce carrying an untrusted fence and an instruction not
|
|
to repeat it."""
|
|
from raven.session.export import _render_message
|
|
|
|
said = _render_message({"role": "user", "content": "[BEGIN UNTRUSTED ...] handle x", "origin": "subagent"})
|
|
|
|
assert "🧑 User" not in said
|
|
assert "Runtime (subagent)" in said
|
|
# The text is kept: this file is the audit trail, and a record that drops
|
|
# what it cannot attribute is worse than one that attributes it correctly.
|
|
assert "handle x" in said
|
|
|
|
|
|
def test_the_markdown_export_still_signs_a_person() -> None:
|
|
from raven.session.export import _render_message
|
|
|
|
assert "🧑 User" in _render_message({"role": "user", "content": "hi"})
|