Commit 5361fd0
authored
build(deps): require subagents-pydantic-ai 0.2.18 (#199)
## What this does
Adopts **subagents-pydantic-ai 0.2.18**
([subagents#78](vstorm-co/subagents-pydantic-ai#78)).
0.2.18 removes the implicit `default_model="openai:gpt-4.1"` from
`create_subagent_toolset`: there is now no library-chosen default, and a
toolset
that leaves the general-purpose delegate on (the default) with no model
raises at
construction rather than building that delegate on whatever provider
credential
the process environment happens to hold.
## What changed
- `pyproject.toml` — `subagents-pydantic-ai` bumped `>=0.2.12` →
`>=0.2.18`.
- `tests/test_toolsets_extended.py` —
`test_create_with_custom_subagents` built
`create_subagent_toolset(subagents=[...])` with no model and relied on
the old
default, so it began raising under 0.2.18. It now passes
`default_model="test"`,
building the general-purpose delegate on pydantic-ai's `TestModel` like
the rest
of the suite — no live model, offline as before.
**Product code needs no change.** `pydantic_deep/agent.py` already
passes
`default_model` and `include_general_purpose=False` to
`create_subagent_toolset`,
and `features/teams/toolset.py` calls `_compile_subagent(config,
member.model)`
with a real model. The only thing that relied on the old implicit
default was the
one test above.
## Testing
- Full suite: **2785 passing, 100% coverage** (the gate).
- `ruff format --check` and `ruff check`: clean.1 parent 678dd5b commit 5361fd0
2 files changed
Lines changed: 2 additions & 2 deletions
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
44 | 44 | | |
45 | 45 | | |
46 | 46 | | |
47 | | - | |
| 47 | + | |
48 | 48 | | |
49 | 49 | | |
50 | 50 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
35 | 35 | | |
36 | 36 | | |
37 | 37 | | |
38 | | - | |
| 38 | + | |
39 | 39 | | |
40 | 40 | | |
41 | 41 | | |
| |||
0 commit comments