commoncompute 0.3.0__tar.gz → 0.3.1__tar.gz

{commoncompute-0.3.0 → commoncompute-0.3.1}/.gitignore

@@ -3,6 +3,7 @@ node_modules/
 dist/
 coverage/
 .wrangler/
+.wrangler.drifted-backup-*/
 .pytest_cache/
 __pycache__/
 *.pyc

{commoncompute-0.3.0 → commoncompute-0.3.1}/LICENSE

@@ -1,6 +1,6 @@
 MIT License
 
-Copyright (c) 2026 Common Compute, Inc.
+Copyright (c) 2026 Common Compute LLC
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

{commoncompute-0.3.0 → commoncompute-0.3.1}/PKG-INFO

@@ -1,10 +1,10 @@
 Metadata-Version: 2.4
 Name: commoncompute
-Version: 0.3.0
+Version: 0.3.1
 Summary: Official Python SDK for Common Compute — the batch AI bill you shouldn't be paying.
 Project-URL: Homepage, https://commoncompute.ai
 Project-URL: Documentation, https://commoncompute.ai/docs
-Project-URL: Source, https://github.com/Ikaikaalika/commoncomputeai
+Project-URL: Source, https://github.com/commoncompute/sdk
 Project-URL: Changelog, https://commoncompute.ai/changelog
 Author-email: Common Compute <support@commoncompute.ai>
 License: MIT

{commoncompute-0.3.0 → commoncompute-0.3.1}/commoncompute/__init__.py

@@ -30,7 +30,7 @@ in the catalog — the API declines them with `workload_not_available`
 # `commoncompute.__version__` at import time for the User-Agent header,
 # and the previous ordering (`__version__` defined at the bottom) made
 # every clean install of the SDK fail with a circular ImportError.
-__version__ = "0.3.0"
+__version__ = "0.3.1"
 
 from .client import Client, AsyncClient
 from .errors import (

{commoncompute-0.3.0 → commoncompute-0.3.1}/commoncompute/_async_client.py

@@ -278,8 +278,10 @@ class _KeysAsync:
     async def list(self) -> list[dict[str, Any]]:
         return (await self._t.request("GET", "/v1/me/keys")).get("keys", [])
 
-    async def create(self, *, name: str, mode: str = "live", scopes: "Optional[Sequence[str]]" = None) -> dict[str, Any]:
+    async def create(self, *, name: str, mode: str = "live", scopes: "Optional[Sequence[str]]" = None, password: "Optional[str]" = None) -> dict[str, Any]:
         body: dict[str, Any] = {"name": name, "mode": mode}
+        if password:
+            body["password"] = password
         if scopes:
             body["scopes"] = scopes
         return await self._t.request("POST", "/v1/me/keys", json_body=body)

{commoncompute-0.3.0 → commoncompute-0.3.1}/commoncompute/_generated_workloads.py

@@ -1,7 +1,10 @@
-# @generated by packages/workloads/codegen/emit-sdk-workloads.ts — DO NOT EDIT.
-# Regenerate:  npx tsx packages/workloads/codegen/emit-sdk-workloads.ts
-# Source of truth: packages/workloads (the shared catalog). CI fails if this
-# file drifts from the catalog (see tests/unit/workloads-parity.test.ts).
+# @generated from Common Compute's internal workload catalog — DO NOT EDIT.
+#
+# This file is emitted by an internal codegen step and is not regenerable from
+# this repository. Edits here will be overwritten; the catalog is the source of
+# truth and drift is caught before release.
+#
+# Contains workload IDs and image-name routing heuristics only.
 from __future__ import annotations
 
 # Every workload id in the catalog. Used to decide whether a jobDefinition head

{commoncompute-0.3.0 → commoncompute-0.3.1}/commoncompute/aws_batch.py

@@ -103,11 +103,10 @@ def _to_iso(ms: Optional[int]) -> Optional[str]:
 # Heuristic: anything containing one of these workload ids gets routed
 # directly. Otherwise we default to ``coreml_embed`` (the most common
 # batch workload migrated from AWS Batch + bge-* embeddings pipelines).
-# Workload identity comes from the generated catalog mirror so this shim
-# resolves jobDefinitions identically to the wire shim (apps/aws-shim) and the
-# Node SDK. The previous hand-maintained set had drifted to 12 of the catalog's
-# 24 workloads. Regenerate with:
-#   npx tsx packages/workloads/codegen/emit-sdk-workloads.ts
+# Workload identity comes from the generated catalog mirror
+# (``_generated_workloads.py``) so this shim resolves jobDefinitions
+# identically to the server side and to the Node SDK, instead of drifting
+# from a hand-maintained list.
 from ._generated_workloads import IMAGE_PATTERNS as _IMAGE_PATTERN_SOURCES
 from ._generated_workloads import KNOWN_WORKLOAD_IDS as _KNOWN_WORKLOADS
 
@@ -403,7 +402,7 @@ class BatchClient:
         """Mirror AWS describe_jobs — returns ``{"jobs": [...]}``.
 
         Just-submitted jobs may not be queryable for ~100ms while the
-        underlying D1 row propagates across the worker's read replicas.
+        underlying row propagates across the backing datastore's read replicas.
         We absorb a single 404 by retrying once after 250ms, which is
         enough to mask the eventual-consistency window in practice."""
         out: list[dict[str, Any]] = []

{commoncompute-0.3.0 → commoncompute-0.3.1}/commoncompute/cli.py

@@ -435,10 +435,46 @@ def _wait_job(client: Client, job_id: str, timeout: float, *, quiet: bool):
         return client.wait(job_id, timeout=timeout, on_progress=_tick)
 
 
+def _parse_payload_arg(payload: str) -> object:
+    """Resolve --payload: @- (stdin), @path or bare path (file), or inline JSON.
+
+    The help has always said "path to JSON payload file", but the code only
+    honoured that with an @ prefix — a bare path fell through to json.loads()
+    and died with a raw JSONDecodeError traceback. Accept the documented
+    form, keep both @ forms, and turn every failure into a clean CLI error.
+    """
+    try:
+        if payload == "@-":
+            return json.load(sys.stdin)
+        if payload.startswith("@"):
+            return json.loads(Path(payload[1:]).read_text())
+        p = Path(payload)
+        if p.exists():
+            return json.loads(p.read_text())
+        # Not a file on disk — treat as inline JSON. A path-like string that
+        # simply doesn't exist gets a hint instead of a JSON parse error.
+        try:
+            return json.loads(payload)
+        except json.JSONDecodeError:
+            if payload.endswith(".json") or "/" in payload:
+                err_console.print(f"[red]Error:[/red] payload file not found: {payload}")
+            else:
+                err_console.print(f"[red]Error:[/red] --payload is not a file, not valid inline JSON: {payload!r}")
+            raise typer.Exit(2)
+    except typer.Exit:
+        raise
+    except json.JSONDecodeError as e:
+        err_console.print(f"[red]Error:[/red] invalid JSON in payload ({e.msg} at line {e.lineno}, column {e.colno})")
+        raise typer.Exit(2)
+    except OSError as e:
+        err_console.print(f"[red]Error:[/red] cannot read payload: {e}")
+        raise typer.Exit(2)
+
+
 @app.command()
 def submit(
     workload: str = typer.Argument(..., help="Workload id."),
-    payload: Optional[str] = typer.Option(None, "--payload", "-f", help="Path to JSON payload file, or @- for stdin."),
+    payload: Optional[str] = typer.Option(None, "--payload", "-f", help="Path to a JSON payload file, @- for stdin, or an inline JSON string."),
     model: Optional[str] = typer.Option(None, "--model", "-m"),
     priority: str = typer.Option("standard", "--priority", "-p"),
     wait: bool = typer.Option(False, "--wait", help="Block until the job finishes."),
@@ -448,12 +484,7 @@ def submit(
     """Submit a native Common Compute job."""
     body: object = None
     if payload:
-        if payload == "@-":
-            body = json.load(sys.stdin)
-        elif payload.startswith("@"):
-            body = json.loads(Path(payload[1:]).read_text())
-        else:
-            body = json.loads(payload)
+        body = _parse_payload_arg(payload)
     c = _client()
     try:
         job = c.jobs.submit(workload, payload=body, model_id=model, priority=priority)

{commoncompute-0.3.0 → commoncompute-0.3.1}/commoncompute/client.py

@@ -263,10 +263,28 @@ class _KeysSync:
     def list(self) -> list[dict[str, Any]]:
         return self._t.request("GET", "/v1/me/keys").get("keys", [])
 
-    def create(self, *, name: str, mode: str = "live", scopes: "Optional[Sequence[str]]" = None) -> dict[str, Any]:
+    def create(
+        self,
+        *,
+        name: str,
+        mode: str = "live",
+        scopes: "Optional[Sequence[str]]" = None,
+        password: "Optional[str]" = None,
+    ) -> dict[str, Any]:
+        """Create an API key.
+
+        Requires a signed-in *session* plus the account password. An API key
+        cannot create another API key (the server returns 403
+        ``api_key_cannot_mint``) — otherwise revoking a leaked key would not
+        revoke the keys it had spawned. Since this SDK normally authenticates
+        with an API key, most callers should create keys in the dashboard at
+        /app/api-keys instead.
+        """
         body: dict[str, Any] = {"name": name, "mode": mode}
         if scopes:
             body["scopes"] = scopes
+        if password:
+            body["password"] = password
         return self._t.request("POST", "/v1/me/keys", json_body=body)
 
     def revoke(self, key_id: str) -> None:

{commoncompute-0.3.0 → commoncompute-0.3.1}/commoncompute/tasks.py

@@ -27,12 +27,31 @@ from .models import Job, Quote, Receipt, TaskResult
 
 FileInput = Union[str, Path, bytes, io.BufferedReader]
 
+# Schemes recognised as "this is a URI, not a local path". Only https:// can
+# actually be fetched: the provider's download gate hard-rejects every other
+# scheme (TrustedDownload.swift — "only https"), and http:// specifically is
+# refused because a job input must not travel in the clear.
+#
+# The others stay in this tuple ON PURPOSE. Dropping them would make an s3://
+# URI fall through to the local-path branch and raise a confusing
+# "No such file: s3://bucket/key"; keeping them lets _file_payload reject the
+# scheme by name, at submit time, instead of the job dispatching and failing
+# on someone's Mac minutes later (2026-08-01 audit — the SDKs documented all
+# five schemes and only one ever worked).
 _URI_SCHEMES = ("http://", "https://", "s3://", "r2://", "gs://")
+_FETCHABLE_URI_SCHEMES = ("https://",)
 
 
 def _file_payload(f: FileInput, *, allowed_suffixes: Optional[Sequence[str]] = None) -> dict:
     """Normalise a file input into the job-payload contract."""
     if isinstance(f, str) and f.startswith(_URI_SCHEMES):
+        if not f.startswith(_FETCHABLE_URI_SCHEMES):
+            scheme = f.split("://", 1)[0]
+            raise UnsupportedFormatError(
+                f"{scheme}:// inputs are not supported — providers can only fetch https:// URLs. "
+                f"Pass a local path (the file is uploaded for you) or a presigned https:// URL.",
+                code="unsupported_input_scheme",
+            )
         return {"input_uri": f}
     if isinstance(f, (str, Path)):
         p = Path(f)
@@ -367,6 +386,14 @@ class Video(_TaskBase):
         max_spend_usd: Optional[float] = None,
         dry_run: bool = False,
     ) -> Union[Job, Quote]:
+        """Transcode a video with VideoToolbox hardware encoders.
+
+        ``output_format`` selects the codec (``"h264"`` / ``"hevc"`` /
+        ``"prores"``); ``codec=`` overrides it if both are given.
+        ``resolution`` is not yet implemented server-side — a job that sets
+        it fails fast with a clear error instead of silently encoding at
+        the source's native size.
+        """
         workload_id, payload, units, model_id = _transcode_request(
             input, output_format, resolution=resolution, codec=codec, output_minutes=output_minutes,
         )

{commoncompute-0.3.0 → commoncompute-0.3.1}/pyproject.toml

@@ -38,7 +38,7 @@ dependencies = [
 [project.urls]
 Homepage = "https://commoncompute.ai"
 Documentation = "https://commoncompute.ai/docs"
-Source = "https://github.com/Ikaikaalika/commoncomputeai"
+Source = "https://github.com/commoncompute/sdk"
 Changelog = "https://commoncompute.ai/changelog"
 
 [project.scripts]