claude-subprocess 0.1.2__tar.gz → 0.1.4__tar.gz

{claude_subprocess-0.1.2 → claude_subprocess-0.1.4}/.gitignore

@@ -1,6 +1,7 @@
 __pycache__/
 *.py[co]
 .pytest_cache/
+.coverage
 .venv/
 *.egg-info/
 graphify-out/

claude_subprocess-0.1.2/README.md → claude_subprocess-0.1.4/PKG-INFO

@@ -1,5 +1,28 @@
+Metadata-Version: 2.5
+Name: claude-subprocess
+Version: 0.1.4
+Summary: Hardened wrapper around the local `claude -p` CLI (subscription auth), + JSON extraction and response cache
+Project-URL: Homepage, https://github.com/octrow/claude-cli
+Project-URL: Issues, https://github.com/octrow/claude-cli/issues
+Author: Daniil Petrov
+License-Expression: MIT
+License-File: LICENSE
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Requires-Python: >=3.12
+Provides-Extra: dev
+Requires-Dist: pytest-asyncio>=0.24; extra == 'dev'
+Requires-Dist: pytest>=8; extra == 'dev'
+Requires-Dist: ruff==0.17.3; extra == 'dev'
+Requires-Dist: ty==0.0.95; extra == 'dev'
+Description-Content-Type: text/markdown
+
 <p align="center">
   <picture>
+    <source media="(max-width: 600px) and (prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/hero-mobile-dark.svg">
+    <source media="(max-width: 600px)" srcset="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/hero-mobile-light.svg">
     <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/hero-dark.svg">
     <img src="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/hero-light.svg" width="100%" alt="claude-subprocess: call claude -p from Python on your Claude Code subscription, with the risky flags locked down. Beside the title, the real argv build_args(model='haiku') produces: omniroute run claude -- -p --model haiku --effort low --output-format json --no-session-persistence --tools '' --disable-slash-commands --setting-sources user --strict-mcp-config --mcp-config mcp-empty.json. --bare is never sent because it breaks OAuth.">
   </picture>
@@ -28,17 +51,17 @@ uv add claude-subprocess   # or: pip install claude-subprocess
 The PyPI name is `claude-subprocess`, the import name is `claude_cli`. At runtime you need
 `omniroute` and `claude` on `PATH`, with Claude Code logged in.
 
-## A call
+## Usage
 
 ```python
 from claude_cli import run_claude, extract_json
 
 res = run_claude('Reply with {"ok": true}', model="haiku", effort="low", timeout=60)
-res.text         # the model's answer
-res.cost_usd     # cost the envelope reported (may be None)
+res.text  # the model's answer
+res.cost_usd  # cost the envelope reported (may be None)
 res.duration_ms
-res.raw          # the full envelope dict
-extract_json(res.text)   # {'ok': True}
+res.raw  # the full envelope dict
+extract_json(res.text)  # {'ok': True}
 ```
 
 `extract_json` tolerates code fences and prose around the object:
@@ -51,6 +74,8 @@ extract_json(res.text)   # {'ok': True}
 ## How a call is routed
 
 <picture>
+  <source media="(max-width: 600px) and (prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/flow-mobile-dark.svg">
+  <source media="(max-width: 600px)" srcset="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/flow-mobile-light.svg">
   <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/flow-dark.svg">
   <img src="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/flow-light.svg" width="100%" alt="run_claude, arun_claude and stream_claude send tool-less calls to OmniRoute's HTTP API at /v1/chat/completions. Calls with tools, a project cwd or stream-json go through omniroute run claude -- -p with a throwaway cwd and an empty MCP config. An HTTP error falls back to the CLI and parks HTTP for 600 seconds. Both paths return a ClaudeResult. NotLoggedInError and UsageLimitError fail fast and derive from ClaudeCliError(RuntimeError).">
 </picture>
@@ -128,9 +153,12 @@ with its `CLAUDE.md` and settings loaded, and lets it write files:
 
 ```python
 res = stream_claude(
-    "/apply data/export/interested/acme.md", on_event,
-    cwd="~/projects/my-repo", slash_commands=True,
-    setting_sources="user,project", permission_mode="acceptEdits",
+    "/apply data/export/interested/acme.md",
+    on_event,
+    cwd="~/projects/my-repo",
+    slash_commands=True,
+    setting_sources="user,project",
+    permission_mode="acceptEdits",
     tools="Read,Write,Edit,Bash,WebFetch",
 )
 ```
@@ -149,8 +177,7 @@ Cached:
 from claude_cli import FileCacheStore, cached_run, run_claude
 
 store = FileCacheStore("~/.cache/myapp/llm.db")
-text = cached_run(run_claude, store, prompt, model="haiku",
-                  purpose="classify", version="v3")
+text = cached_run(run_claude, store, prompt, model="haiku", purpose="classify", version="v3")
 ```
 
 `cache_key(model, prompt, schema=..., version=..., purpose=..., tools=...)` is a sha256 over
@@ -175,14 +202,21 @@ the gateway's model-to-combo mappings, or you pass a combo name directly. `tier=
 res = run_claude(prompt, model="sub-first")  # explicit combo
 ```
 
+`direct=True` skips OmniRoute: `claude -p` runs on the account the local Claude Code is
+logged in to, `model` is a plain Claude model id, and a usage limit ends the session.
+
+```python
+res = run_claude(prompt, model="claude-sonnet-5-5", direct=True)
+```
+
 ## Checks before you spend quota
 
 ```python
 from claude_cli import health, doctor, quota_status, simulate, dry_run_plan
 
-health()                          # omniroute health
-simulate("hi", combo="free-first")
-dry_run_plan()                    # planned command, args and env key *names* only
+health()  # omniroute health
+simulate("hi", combo="free-stack")
+dry_run_plan()  # planned command, args and env key *names* only
 ```
 
 None of these raise or spend quota; each returns `(ok, text)`.

claude_subprocess-0.1.2/PKG-INFO → claude_subprocess-0.1.4/README.md

@@ -1,20 +1,7 @@
-Metadata-Version: 2.5
-Name: claude-subprocess
-Version: 0.1.2
-Summary: Hardened wrapper around the local `claude -p` CLI (subscription auth), + JSON extraction and response cache
-Project-URL: Homepage, https://github.com/octrow/claude-cli
-Project-URL: Issues, https://github.com/octrow/claude-cli/issues
-Author: Daniil Petrov
-License-Expression: MIT
-License-File: LICENSE
-Requires-Python: >=3.12
-Provides-Extra: dev
-Requires-Dist: pytest-asyncio>=0.24; extra == 'dev'
-Requires-Dist: pytest>=8; extra == 'dev'
-Description-Content-Type: text/markdown
-
 <p align="center">
   <picture>
+    <source media="(max-width: 600px) and (prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/hero-mobile-dark.svg">
+    <source media="(max-width: 600px)" srcset="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/hero-mobile-light.svg">
     <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/hero-dark.svg">
     <img src="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/hero-light.svg" width="100%" alt="claude-subprocess: call claude -p from Python on your Claude Code subscription, with the risky flags locked down. Beside the title, the real argv build_args(model='haiku') produces: omniroute run claude -- -p --model haiku --effort low --output-format json --no-session-persistence --tools '' --disable-slash-commands --setting-sources user --strict-mcp-config --mcp-config mcp-empty.json. --bare is never sent because it breaks OAuth.">
   </picture>
@@ -43,17 +30,17 @@ uv add claude-subprocess   # or: pip install claude-subprocess
 The PyPI name is `claude-subprocess`, the import name is `claude_cli`. At runtime you need
 `omniroute` and `claude` on `PATH`, with Claude Code logged in.
 
-## A call
+## Usage
 
 ```python
 from claude_cli import run_claude, extract_json
 
 res = run_claude('Reply with {"ok": true}', model="haiku", effort="low", timeout=60)
-res.text         # the model's answer
-res.cost_usd     # cost the envelope reported (may be None)
+res.text  # the model's answer
+res.cost_usd  # cost the envelope reported (may be None)
 res.duration_ms
-res.raw          # the full envelope dict
-extract_json(res.text)   # {'ok': True}
+res.raw  # the full envelope dict
+extract_json(res.text)  # {'ok': True}
 ```
 
 `extract_json` tolerates code fences and prose around the object:
@@ -66,6 +53,8 @@ extract_json(res.text)   # {'ok': True}
 ## How a call is routed
 
 <picture>
+  <source media="(max-width: 600px) and (prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/flow-mobile-dark.svg">
+  <source media="(max-width: 600px)" srcset="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/flow-mobile-light.svg">
   <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/flow-dark.svg">
   <img src="https://raw.githubusercontent.com/octrow/claude-cli/main/assets/readme/flow-light.svg" width="100%" alt="run_claude, arun_claude and stream_claude send tool-less calls to OmniRoute's HTTP API at /v1/chat/completions. Calls with tools, a project cwd or stream-json go through omniroute run claude -- -p with a throwaway cwd and an empty MCP config. An HTTP error falls back to the CLI and parks HTTP for 600 seconds. Both paths return a ClaudeResult. NotLoggedInError and UsageLimitError fail fast and derive from ClaudeCliError(RuntimeError).">
 </picture>
@@ -143,9 +132,12 @@ with its `CLAUDE.md` and settings loaded, and lets it write files:
 
 ```python
 res = stream_claude(
-    "/apply data/export/interested/acme.md", on_event,
-    cwd="~/projects/my-repo", slash_commands=True,
-    setting_sources="user,project", permission_mode="acceptEdits",
+    "/apply data/export/interested/acme.md",
+    on_event,
+    cwd="~/projects/my-repo",
+    slash_commands=True,
+    setting_sources="user,project",
+    permission_mode="acceptEdits",
     tools="Read,Write,Edit,Bash,WebFetch",
 )
 ```
@@ -164,8 +156,7 @@ Cached:
 from claude_cli import FileCacheStore, cached_run, run_claude
 
 store = FileCacheStore("~/.cache/myapp/llm.db")
-text = cached_run(run_claude, store, prompt, model="haiku",
-                  purpose="classify", version="v3")
+text = cached_run(run_claude, store, prompt, model="haiku", purpose="classify", version="v3")
 ```
 
 `cache_key(model, prompt, schema=..., version=..., purpose=..., tools=...)` is a sha256 over
@@ -190,14 +181,21 @@ the gateway's model-to-combo mappings, or you pass a combo name directly. `tier=
 res = run_claude(prompt, model="sub-first")  # explicit combo
 ```
 
+`direct=True` skips OmniRoute: `claude -p` runs on the account the local Claude Code is
+logged in to, `model` is a plain Claude model id, and a usage limit ends the session.
+
+```python
+res = run_claude(prompt, model="claude-sonnet-5-5", direct=True)
+```
+
 ## Checks before you spend quota
 
 ```python
 from claude_cli import health, doctor, quota_status, simulate, dry_run_plan
 
-health()                          # omniroute health
-simulate("hi", combo="free-first")
-dry_run_plan()                    # planned command, args and env key *names* only
+health()  # omniroute health
+simulate("hi", combo="free-stack")
+dry_run_plan()  # planned command, args and env key *names* only
 ```
 
 None of these raise or spend quota; each returns `(ok, text)`.

{claude_subprocess-0.1.2 → claude_subprocess-0.1.4}/pyproject.toml

@@ -1,11 +1,17 @@
 [project]
 name = "claude-subprocess"
-version = "0.1.2"
+version = "0.1.4"
 license = "MIT"
 readme = "README.md"
 authors = [{ name = "Daniil Petrov" }]
 description = "Hardened wrapper around the local `claude -p` CLI (subscription auth), + JSON extraction and response cache"
 requires-python = ">=3.12"
+classifiers = [
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13",
+    "Operating System :: OS Independent",
+]
 dependencies = []  # stdlib only, on purpose
 
 [project.urls]
@@ -13,7 +19,7 @@ Homepage = "https://github.com/octrow/claude-cli"
 Issues = "https://github.com/octrow/claude-cli/issues"
 
 [project.optional-dependencies]
-dev = ["pytest>=8", "pytest-asyncio>=0.24"]
+dev = ["pytest>=8", "pytest-asyncio>=0.24", "ruff==0.17.3", "ty==0.0.95"]
 
 [build-system]
 requires = ["hatchling"]
@@ -28,3 +34,11 @@ include = ["src", "tests", "README.md", "LICENSE"]
 [tool.pytest.ini_options]
 testpaths = ["tests"]
 asyncio_mode = "auto"
+
+[tool.ruff]
+line-length = 100  # the code was already written to ~100; only 6 lines were longer
+
+[tool.vulture]
+paths = ["src", "tests"]
+min_confidence = 80
+ignore_names = ["auto"]  # pytest fixture requested for its env side effect, never read

{claude_subprocess-0.1.2 → claude_subprocess-0.1.4}/src/claude_cli/cache.py

@@ -12,7 +12,7 @@ import hashlib
 import logging
 import sqlite3
 from contextlib import closing
-from datetime import datetime, timedelta, timezone
+from datetime import UTC, datetime, timedelta
 from pathlib import Path
 from typing import Protocol
 
@@ -116,8 +116,8 @@ class FileCacheStore:
         except ValueError:  # unparseable timestamp: treat as expired, re-ask rather than trust
             return False
         if written.tzinfo is None:  # naive row (hand-written/legacy): assume UTC, never TypeError
-            written = written.replace(tzinfo=timezone.utc)
-        age = datetime.now(timezone.utc) - written
+            written = written.replace(tzinfo=UTC)
+        age = datetime.now(UTC) - written
         return age <= timedelta(days=self.max_age_days)
 
     def put(self, key: str, text: str, **meta) -> None:
@@ -134,7 +134,7 @@ class FileCacheStore:
                     text,
                     meta.get("cost_usd"),
                     meta.get("duration_ms"),
-                    datetime.now(timezone.utc).isoformat(),
+                    datetime.now(UTC).isoformat(),
                 ),
             )
             conn.commit()
@@ -186,8 +186,12 @@ def cached_run(
     # folded into the model part only when set, so existing rows stay valid.
     routing = ",".join(f"{k}={kwargs[k]}" for k in _ROUTING_KWARGS if kwargs.get(k))
     key = cache_key(
-        f"{model}|{routing}" if routing else model, prompt, schema=schema,
-        version=version, purpose=purpose, tools=kwargs.get("tools", ""),
+        f"{model}|{routing}" if routing else model,
+        prompt,
+        schema=schema,
+        version=version,
+        purpose=purpose,
+        tools=kwargs.get("tools", ""),
     )
     if not refresh:
         hit = store.get(key)
@@ -197,9 +201,13 @@ def cached_run(
     result = runner(prompt, model=model, **kwargs)
     try:
         store.put(
-            key, result.text, model=model, purpose=purpose,
+            key,
+            result.text,
+            model=model,
+            purpose=purpose,
             prompt=prompt if store_prompt else None,
-            cost_usd=result.cost_usd, duration_ms=result.duration_ms,
+            cost_usd=result.cost_usd,
+            duration_ms=result.duration_ms,
         )
     except (sqlite3.Error, OSError) as exc:
         log.warning("llm cache write failed: %s", exc)

{claude_subprocess-0.1.2 → claude_subprocess-0.1.4}/src/claude_cli/http_api.py

@@ -26,7 +26,9 @@ COMBOS = {"haiku": "free-stack", "sonnet": "free-stack", "opus": "hard"}
 COOLDOWN_SECS = 600
 ENV_FILE = Path.home() / ".omniroute" / ".env"
 
-_down_until = 0.0  # ponytail: per process, not per model; a dead gateway costs one timeout per window
+_down_until = (
+    0.0  # ponytail: per process, not per model; a dead gateway costs one timeout per window
+)
 
 
 class HttpLlmError(ClaudeCliError):
@@ -47,8 +49,14 @@ def api_key() -> str | None:
     return None
 
 
-def omniroute_chat(prompt: str, *, model: str = "sonnet", timeout: float = 300,
-                   system_prompt: str | None = None, **_ignored) -> ClaudeResult:
+def omniroute_chat(
+    prompt: str,
+    *,
+    model: str = "sonnet",
+    timeout: float = 300,
+    system_prompt: str | None = None,
+    **_ignored,
+) -> ClaudeResult:
     """One chat completion, ClaudeResult-shaped. CLI-only knobs (effort, cwd...) are ignored."""
     key = api_key()
     if not key:
@@ -58,8 +66,9 @@ def omniroute_chat(prompt: str, *, model: str = "sonnet", timeout: float = 300,
     if system_prompt:
         messages.insert(0, {"role": "system", "content": system_prompt})
     body = json.dumps({"model": COMBOS.get(model, model), "messages": messages}).encode()
-    req = urllib.request.Request(url, body, {
-        "Authorization": f"Bearer {key}", "Content-Type": "application/json"})
+    req = urllib.request.Request(
+        url, body, {"Authorization": f"Bearer {key}", "Content-Type": "application/json"}
+    )
     t0 = time.monotonic()
     try:
         with urllib.request.urlopen(req, timeout=timeout) as resp:
@@ -73,10 +82,16 @@ def omniroute_chat(prompt: str, *, model: str = "sonnet", timeout: float = 300,
         raise HttpLlmError("empty completion")
     # raw keeps the OpenAI body, but `usage` in the CLI envelope's keys so meters read it
     usage = data.get("usage") or {}
-    raw = {**data, "usage": {"input_tokens": usage.get("prompt_tokens"),
-                             "output_tokens": usage.get("completion_tokens")}}
-    return ClaudeResult(text=text, cost_usd=0.0, raw=raw,
-                        duration_ms=int((time.monotonic() - t0) * 1000))
+    raw = {
+        **data,
+        "usage": {
+            "input_tokens": usage.get("prompt_tokens"),
+            "output_tokens": usage.get("completion_tokens"),
+        },
+    }
+    return ClaudeResult(
+        text=text, cost_usd=0.0, raw=raw, duration_ms=int((time.monotonic() - t0) * 1000)
+    )
 
 
 def try_http(prompt: str, *, backend: str, cli_only: bool, **kw) -> ClaudeResult | None:
@@ -97,6 +112,7 @@ def try_http(prompt: str, *, backend: str, cli_only: bool, **kw) -> ClaudeResult
         return omniroute_chat(prompt, **kw)
     except HttpLlmError as exc:
         _down_until = time.monotonic() + COOLDOWN_SECS
-        log.warning("omniroute http failed (%s); CLI fallback, http off for %ds",
-                    exc, COOLDOWN_SECS)
+        log.warning(
+            "omniroute http failed (%s); CLI fallback, http off for %ds", exc, COOLDOWN_SECS
+        )
         return None

{claude_subprocess-0.1.2 → claude_subprocess-0.1.4}/src/claude_cli/parsing.py

@@ -13,8 +13,15 @@ from typing import Any
 
 #: Metadata keys lifted from a stream-json ``result`` event (tlf/digest.py).
 _META_KEYS = (
-    "usage", "total_cost_usd", "duration_ms", "duration_api_ms",
-    "num_turns", "session_id", "model", "modelUsage", "is_error",
+    "usage",
+    "total_cost_usd",
+    "duration_ms",
+    "duration_api_ms",
+    "num_turns",
+    "session_id",
+    "model",
+    "modelUsage",
+    "is_error",
 )
 
 _FENCED_JSON_RE = re.compile(r"```(?:json)?\s*(\{.*\})\s*```", re.DOTALL)
@@ -58,25 +65,8 @@ def extract_json(text: str) -> Any:
     start = text.find("{")
     if start == -1:
         return json.loads(text)
-    depth, in_str, esc = 0, False, False
-    for i in range(start, len(text)):
-        ch = text[i]
-        if in_str:
-            if esc:
-                esc = False
-            elif ch == "\\":
-                esc = True
-            elif ch == '"':
-                in_str = False
-        elif ch == '"':
-            in_str = True
-        elif ch == "{":
-            depth += 1
-        elif ch == "}":
-            depth -= 1
-            if depth == 0:
-                return json.loads(text[start : i + 1])
-    return json.loads(text[start:])  # unbalanced — let json raise a clear error
+    # raw_decode parses the first object at `start` and ignores what follows it
+    return json.JSONDecoder().raw_decode(text, start)[0]
 
 
 def extract_json_or_none(text: str) -> Any | None:
@@ -110,9 +100,7 @@ def _is_omniroute_banner_line(line: str) -> bool:
         return True
     if " is ignored, " in line and "set it first" in line:
         return True
-    if "connectors are disabled because ANTHROPIC_API_KEY" in line:
-        return True
-    return False
+    return "connectors are disabled because ANTHROPIC_API_KEY" in line
 
 
 def strip_omniroute_noise(text: str) -> str:
@@ -151,7 +139,8 @@ def parse_envelope(stdout: str) -> dict:
         if env is None:
             raise ValueError(f"claude returned non-JSON: {stdout[-300:]}") from exc
     if not isinstance(env, dict):
-        raise ValueError(f"unexpected claude envelope: {type(env).__name__}")
+        # ValueError, not TypeError: runner._finish catches ValueError for every bad envelope
+        raise ValueError(f"unexpected claude envelope: {type(env).__name__}")  # noqa: TRY004
     return env