feat(my-deepagent): v0.1.0 Step 0~5 — scaffolding through deepagent + OpenRouter

Python rewrite of the agent harness on top of deepagents 0.6.1 + langchain 1.x,
replacing the abandoned TS attempt in packages/. 388 unit/integration tests pass.

Steps
-----
0. Scaffolding — uv workspace, ruff/mypy/pre-commit/alembic, src/tests/docs
   trees with docs/schemas/ seeded from my-deepagent-seed/.
1. Core — config (pydantic-settings with MYDEEPAGENT_ env prefix and TOML
   source), enums (Backend, Capability, RiskLevel, ApprovalDecisionAction,
   ApprovalState, RunState, RunPhaseState, SessionState, ErrorClass),
   errors (MyDeepAgentError + BudgetExhaustedError with PEP-3134 cause +
   context suppression), hash (canonical JSON + sha256).
2. Persona/Workflow/Binding — pydantic v2 schemas with tuple-based deep
   immutability (post-construction hash drift prevented), YAML loaders,
   deterministic auto-select (preferred_backends → version → name → hash),
   override resolution with ineligibility diagnostics, PersonaConsentStore
   with fcntl.flock + tmp+fsync+rename atomic write.
3. Artifact schema registry — Draft202012Validator, multi-root resolution,
   structured ValidationFinding output.
4. Persistence — 18 SQLAlchemy 2.0 async ORM models with FK CASCADE/RESTRICT,
   WAL + busy_timeout + foreign_keys PRAGMA, alembic baseline +
   ux_active_run_repo_base partial unique index, LangGraph SqliteSaver as
   context manager only (lifecycle safety).
5. DeepAgent session — build_agent wires Persona → create_deep_agent with
   LocalShellBackend / FilesystemBackend / StateBackend / CompositeBackend,
   ChatOpenAI(base_url=openrouter) for openrouter: model strings, and 4
   middleware classes (cost / audit-tool / safety-shell / fallback-model).

Critical workarounds
--------------------
- deepagents 0.6.1 rejects FilesystemPermission together with backends that
  implement SandboxBackendProtocol (LocalShellBackend). SafetyShellMiddleware
  enforces destructive-command and secret-path policy at the tool layer
  instead, and build_agent strips the permissions kwarg when the persona's
  deepagents_backend is local_shell.
- FilesystemOperation in deepagents is Literal['read', 'write'] only;
  _map_operations collapses our richer schema (read/write/edit/ls) safely.

Real OpenRouter smoke
---------------------
test_openrouter_deepagents_local_shell_smoke calls DeepSeek via deepagents +
LocalShellBackend + SafetyShellMiddleware end-to-end. PASS, ~$0.000001 cost,
input=9 / output=1 tokens with content "OK".

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
chungyeong
2026-05-15 19:40:02 +09:00
parent 1fe59d16ca
commit 17ba5d723b
100 changed files with 12408 additions and 0 deletions

View File

@@ -0,0 +1,109 @@
"""Application configuration loaded from env, .env, and TOML file via pydantic-settings."""
from __future__ import annotations
from pathlib import Path
from typing import Literal
from platformdirs import PlatformDirs
from pydantic import Field, ValidationError, field_validator
from pydantic_settings import (
BaseSettings,
PydanticBaseSettingsSource,
SettingsConfigDict,
TomlConfigSettingsSource,
)
from .enums import ErrorClass
from .errors import MyDeepAgentError
_DIRS = PlatformDirs("my-deepagent", "user", roaming=False)
class Config(BaseSettings):
"""Frozen application config. Source priority (high -> low): CLI/env, .env, TOML, defaults."""
model_config = SettingsConfigDict(
env_prefix="MYDEEPAGENT_",
env_file=".env",
env_file_encoding="utf-8",
toml_file=Path(_DIRS.user_config_dir) / "config.toml",
frozen=True,
extra="ignore",
)
# storage
database_url: str = Field(
default_factory=lambda: (
f"sqlite+aiosqlite:///{Path(_DIRS.user_data_dir) / 'database.sqlite3'}"
)
)
workspace_root: Path = Field(default_factory=Path.cwd)
data_dir: Path = Field(default_factory=lambda: Path(_DIRS.user_data_dir))
config_dir: Path = Field(default_factory=lambda: Path(_DIRS.user_config_dir))
state_dir: Path = Field(default_factory=lambda: Path(_DIRS.user_state_dir))
# logging / i18n
log_level: Literal["trace", "debug", "info", "warn", "error"] = "info"
lang: Literal["ko", "en"] = "ko"
# providers
openrouter_api_key: str | None = None
openrouter_base_url: str = "https://openrouter.ai/api/v1"
# observability
langsmith_tracing: bool = False
langsmith_api_key: str | None = None
langsmith_project: str = "my-deepagent"
# budget
budget_daily_usd: float = Field(default=5.0, ge=0)
budget_daily_warn_usd: float = Field(default=3.0, ge=0)
budget_run_usd: float = Field(default=1.0, ge=0)
budget_run_warn_usd: float = Field(default=0.5, ge=0)
budget_on_hit: Literal["prompt", "block", "warn_continue"] = "prompt"
# defaults
default_persona: str = "default-interactive"
@field_validator("workspace_root", "data_dir", "config_dir", "state_dir")
@classmethod
def _expand(cls, v: Path) -> Path:
return Path(v).expanduser().resolve()
@classmethod
def settings_customise_sources(
cls,
settings_cls: type[BaseSettings],
init_settings: PydanticBaseSettingsSource,
env_settings: PydanticBaseSettingsSource,
dotenv_settings: PydanticBaseSettingsSource,
file_secret_settings: PydanticBaseSettingsSource,
) -> tuple[PydanticBaseSettingsSource, ...]:
# priority: init > env > dotenv > toml > defaults
return (
init_settings,
env_settings,
dotenv_settings,
TomlConfigSettingsSource(settings_cls),
file_secret_settings,
)
def load_config(**overrides: object) -> Config:
"""Load Config with optional kwargs override.
Wraps pydantic ValidationError in MyDeepAgentError(fatal, config_invalid) per plan §18.
"""
try:
return Config(**overrides) # type: ignore[arg-type]
except ValidationError as e:
raise MyDeepAgentError(
ErrorClass.FATAL,
"config_invalid",
message=f"config validation failed: {e}",
recovery_hint=(
"check .env, environment variables, and ~/.config/my-deepagent/config.toml"
),
cause=e,
) from e