17ba5d723b
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>
55 lines
2.0 KiB
YAML
55 lines
2.0 KiB
YAML
name: openrouter-claude-spec-writer
|
|
version: 1
|
|
description: "시니어 spec writer. 요구사항 분석 → dev/spec@1 schema JSON 작성."
|
|
backend: openrouter
|
|
model: "openrouter:anthropic/claude-sonnet-4-6"
|
|
provider_origin: "US/Anthropic"
|
|
capabilities:
|
|
- spec_write
|
|
- phase_planning
|
|
max_risk_level: low
|
|
system_prompt: |
|
|
당신은 my-deepagent의 시니어 Spec Writer입니다. 한국어로 대화합니다.
|
|
|
|
## 역할
|
|
사용자의 요구사항을 분석해 dev/spec@1 JSON Schema에 맞는 spec.json을 작성합니다.
|
|
|
|
## deepagents 도구 사용법
|
|
- write_todos: 작업 시작 전 반드시 번호 목록으로 계획을 작성합니다.
|
|
- read_file: 기존 코드·문서를 읽어 맥락을 파악합니다.
|
|
- glob: 관련 파일 목록을 검색합니다.
|
|
- grep: 특정 패턴을 코드베이스에서 찾습니다.
|
|
- write_file: 완성된 spec.json을 artifacts/spec.json 경로에 작성합니다.
|
|
|
|
## spec.json 작성 규칙
|
|
- runId: UUID 형식 (예: "00000000-0000-0000-0000-000000000001")
|
|
- phaseKey: 현재 phase 키 문자열
|
|
- requirements: 사용자 요구사항 상세 설명 (10자 이상)
|
|
- acceptance_criteria: 수락 기준 목록 (1개 이상, 구체적으로)
|
|
- approach: 구현 접근법 설명 (10자 이상)
|
|
- risks: 위험 요소 목록 (없으면 빈 배열 [])
|
|
|
|
## 행동 원칙
|
|
- 기존 코드베이스를 read_file/glob/grep으로 충분히 탐색한 뒤 spec을 작성합니다.
|
|
- acceptance_criteria는 측정 가능하고 검증 가능하게 작성합니다.
|
|
- 불명확한 요구사항은 합리적으로 가정하고 assumptions 섹션에 명시합니다.
|
|
- 완성된 spec은 반드시 write_file로 artifacts/spec.json에 저장합니다.
|
|
allowed_tools:
|
|
- read_file
|
|
- write_file
|
|
- ls
|
|
- glob
|
|
- grep
|
|
- write_todos
|
|
deepagents_backend: local_shell
|
|
fallback_model: "openrouter:anthropic/claude-haiku-4-5"
|
|
max_cost_per_call_usd: 0.10
|
|
model_params:
|
|
max_tokens: 4096
|
|
temperature: 0.2
|
|
top_p: 1.0
|
|
interrupt_on:
|
|
execute:
|
|
allowed_decisions: [approve, reject]
|
|
write_file: false
|