sference-cli 0.1.1__tar.gz → 0.1.2__tar.gz

{sference_cli-0.1.1 → sference_cli-0.1.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sference-cli
-Version: 0.1.1
+Version: 0.1.2
 Summary: sference command-line interface
 Project-URL: Homepage, https://sference.com
 Project-URL: Repository, https://github.com/s-ference/sference
@@ -21,7 +21,7 @@ Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
 Classifier: Topic :: Utilities
 Requires-Python: >=3.10
-Requires-Dist: sference-sdk>=0.1.1
+Requires-Dist: sference-sdk>=0.1.2
 Requires-Dist: typer>=0.24.1
 Description-Content-Type: text/markdown
 
@@ -33,7 +33,7 @@ Command-line interface for the sference batch API (`sference`). It uses the Pyth
 
 ```bash
 # One-line install (macOS / Linux)
-curl -fsSL https://raw.githubusercontent.com/s-ference/sference/main/install.sh | sh
+eval "$(curl -fsSL https://raw.githubusercontent.com/s-ference/sference/main/install.sh)"
 ```
 
 Or install from PyPI:
@@ -122,35 +122,33 @@ sference responses tail --stream-id <stream_id>
 | `sference auth login` | Store an API key (optional `--api-key`, `--no-browser`) |
 | `sference auth me` | Show current user (`--json` for machine-readable output) |
 
-### Launch (hidden from `sference --help`)
-
-These commands are intentionally omitted from the top-level CLI help; see this section for usage.
+### Launch
 
 | Command | Description |
 |---------|-------------|
 | `sference launch claude` | Launch Claude Code with Sference API routing (`ANTHROPIC_BASE_URL`, `ANTHROPIC_AUTH_TOKEN`, `ANTHROPIC_MODEL`) |
 | `sference launch claude --dry-run` | Print env + command without starting Claude Code |
-| `sference launch claude --model moonshotai/Kimi-K2.6` | Override catalog model |
+| `sference launch claude --model moonshotai/Kimi-K2.7-Code` | Override catalog model |
 | `sference launch claude --enable-tool-search` | Set `ENABLE_TOOL_SEARCH=true` on custom hosts |
 | `sference launch claude --resume` | Forward Claude Code flags/args after `sference launch claude` |
 
-Requires the [Claude Code CLI](https://docs.anthropic.com/en/docs/claude-code) on `PATH`. Uses `~/.sference/credentials.json` or `SFERENCE_API_KEY`. Default model: `moonshotai/Kimi-K2.6` (override with `--model` or `SFERENCE_MODEL`).
+Requires the [Claude Code CLI](https://docs.anthropic.com/en/docs/claude-code) on `PATH`. Uses `~/.sference/credentials.json` or `SFERENCE_API_KEY`. Default model: `moonshotai/Kimi-K2.7-Code` (override with `--model` or `SFERENCE_MODEL`).
 
 | Command | Description |
 |---------|-------------|
 | `sference launch pi` | Launch Pi with Sference API routing (writes `~/.pi/agent/models.json`) |
 | `sference launch pi --dry-run` | Print provider config and command without launching Pi |
-| `sference launch pi --model moonshotai/Kimi-K2.6` | Override catalog model |
+| `sference launch pi --model moonshotai/Kimi-K2.7-Code` | Override catalog model |
 | `sference launch pi -- /path/to/project` | Forward Pi args after `sference launch pi` |
 
-Requires the [Pi CLI](https://pi.dev/docs) on `PATH`. Writes a `sference` provider block to `~/.pi/agent/models.json` with `baseUrl`, `apiKey`, and the chosen `model`, then execs `pi --provider sference --model <id>`. Uses `~/.sference/credentials.json` or `SFERENCE_API_KEY`. Default model: `moonshotai/Kimi-K2.6` (override with `--model` or `SFERENCE_MODEL`).
+Requires the [Pi CLI](https://pi.dev/docs) on `PATH`. Writes a `sference` provider block to `~/.pi/agent/models.json` with `baseUrl`, `apiKey`, and the chosen `model`, then execs `pi --provider sference --model <id>`. Uses `~/.sference/credentials.json` or `SFERENCE_API_KEY`. Default model: `moonshotai/Kimi-K2.7-Code` (override with `--model` or `SFERENCE_MODEL`).
 
 ### Batch
 
 | Command | Description |
 |---------|-------------|
 | `sference batch list` | List batches (table; `--json` for raw payload) |
-| `sference batch submit` | Submit a JSONL file (`--input-file`, optional `--model` for content-only lines, `--window` must be `24h`) |
+| `sference batch submit` | Submit a JSONL file (`--input-file`, optional `--model` for content-only lines, `--window` `15m`/`1h`/`24h`) |
 | `sference batch stream` | Submit, wait, print **JSONL results on stdout** (see below) |
 | `sference batch status` | Get one batch (`--batch-id`, `--json`) |
 | `sference batch wait` | Poll until terminal state (`--batch-id`, `--poll-interval`, `--timeout`, `--json`) |
@@ -172,7 +170,7 @@ Long-lived **streams** are separate from **batches**: you create a stream, submi
 
 | Command | Description |
 |---------|-------------|
-| `sference stream create` | Create a stream (`--name`, `--window` `1h` or `24h`, `--json`) |
+| `sference stream create` | Create a stream (`--name`, `--window` `15m`/`1h`/`24h`, `--json`) |
 | `sference stream list` | List streams (`--json`) |
 | `sference stream status` | Full detail + counters (`--stream-id`, `--json`) |
 | `sference stream submit` | Create responses from JSONL via `POST /v1/responses` per line (`metadata.stream_id` set automatically; `--stream-id`, `--input-file`, `--model` required for content-only lines) — per line: OpenAI batch-style `{custom_id?, method, url, body}` or content-only `{content}` |
@@ -250,9 +248,19 @@ sference batch stream --input-file fixtures/example_batch.jsonl --poll-interval
 
 The SDK and CLI accept two line shapes (see also [`fixtures/example_batch.jsonl`](fixtures/example_batch.jsonl)):
 
-1. **OpenAI-compatible:** each line has `custom_id`, `method`, `url`, and `body` (e.g. chat completions payload with per-line `model`). The CLI `--model` flag is ignored for these lines (a warning may be emitted by the SDK).
+1. **OpenAI-compatible envelope:** each line has `custom_id`, `method`, `url`, and `body`. The API receives only `custom_id` + inner `body` (`method`/`url` are ignored). Inner `body` may use:
+   - **Chat completions:** `messages` (non-empty array), plus optional `temperature`, `max_tokens`, `tools`, …
+   - **Responses API:** `input` (string or message array), optional `instructions`, `max_output_tokens`, … — normalized to chat format at batch create.
 2. **Content-only:** each line is `{"content": "..."}`. Then **`--model` is required** on submit/stream.
 
+Invalid row bodies return **HTTP 400** at create with `requests[i]` and `custom_id` in the error message (nothing is enqueued). Do not set `background: true` inside batch row bodies.
+
+Example Responses-shaped JSONL line:
+
+```json
+{"custom_id":"r2","method":"POST","url":"/v1/responses","body":{"model":"Qwen/Qwen3.6-35B-A3B","input":[{"role":"user","content":"hi"}],"max_output_tokens":256}}
+```
+
 ---
 
 ## Python SDK

{sference_cli-0.1.1 → sference_cli-0.1.2}/README.md

@@ -6,7 +6,7 @@ Command-line interface for the sference batch API (`sference`). It uses the Pyth
 
 ```bash
 # One-line install (macOS / Linux)
-curl -fsSL https://raw.githubusercontent.com/s-ference/sference/main/install.sh | sh
+eval "$(curl -fsSL https://raw.githubusercontent.com/s-ference/sference/main/install.sh)"
 ```
 
 Or install from PyPI:
@@ -95,35 +95,33 @@ sference responses tail --stream-id <stream_id>
 | `sference auth login` | Store an API key (optional `--api-key`, `--no-browser`) |
 | `sference auth me` | Show current user (`--json` for machine-readable output) |
 
-### Launch (hidden from `sference --help`)
-
-These commands are intentionally omitted from the top-level CLI help; see this section for usage.
+### Launch
 
 | Command | Description |
 |---------|-------------|
 | `sference launch claude` | Launch Claude Code with Sference API routing (`ANTHROPIC_BASE_URL`, `ANTHROPIC_AUTH_TOKEN`, `ANTHROPIC_MODEL`) |
 | `sference launch claude --dry-run` | Print env + command without starting Claude Code |
-| `sference launch claude --model moonshotai/Kimi-K2.6` | Override catalog model |
+| `sference launch claude --model moonshotai/Kimi-K2.7-Code` | Override catalog model |
 | `sference launch claude --enable-tool-search` | Set `ENABLE_TOOL_SEARCH=true` on custom hosts |
 | `sference launch claude --resume` | Forward Claude Code flags/args after `sference launch claude` |
 
-Requires the [Claude Code CLI](https://docs.anthropic.com/en/docs/claude-code) on `PATH`. Uses `~/.sference/credentials.json` or `SFERENCE_API_KEY`. Default model: `moonshotai/Kimi-K2.6` (override with `--model` or `SFERENCE_MODEL`).
+Requires the [Claude Code CLI](https://docs.anthropic.com/en/docs/claude-code) on `PATH`. Uses `~/.sference/credentials.json` or `SFERENCE_API_KEY`. Default model: `moonshotai/Kimi-K2.7-Code` (override with `--model` or `SFERENCE_MODEL`).
 
 | Command | Description |
 |---------|-------------|
 | `sference launch pi` | Launch Pi with Sference API routing (writes `~/.pi/agent/models.json`) |
 | `sference launch pi --dry-run` | Print provider config and command without launching Pi |
-| `sference launch pi --model moonshotai/Kimi-K2.6` | Override catalog model |
+| `sference launch pi --model moonshotai/Kimi-K2.7-Code` | Override catalog model |
 | `sference launch pi -- /path/to/project` | Forward Pi args after `sference launch pi` |
 
-Requires the [Pi CLI](https://pi.dev/docs) on `PATH`. Writes a `sference` provider block to `~/.pi/agent/models.json` with `baseUrl`, `apiKey`, and the chosen `model`, then execs `pi --provider sference --model <id>`. Uses `~/.sference/credentials.json` or `SFERENCE_API_KEY`. Default model: `moonshotai/Kimi-K2.6` (override with `--model` or `SFERENCE_MODEL`).
+Requires the [Pi CLI](https://pi.dev/docs) on `PATH`. Writes a `sference` provider block to `~/.pi/agent/models.json` with `baseUrl`, `apiKey`, and the chosen `model`, then execs `pi --provider sference --model <id>`. Uses `~/.sference/credentials.json` or `SFERENCE_API_KEY`. Default model: `moonshotai/Kimi-K2.7-Code` (override with `--model` or `SFERENCE_MODEL`).
 
 ### Batch
 
 | Command | Description |
 |---------|-------------|
 | `sference batch list` | List batches (table; `--json` for raw payload) |
-| `sference batch submit` | Submit a JSONL file (`--input-file`, optional `--model` for content-only lines, `--window` must be `24h`) |
+| `sference batch submit` | Submit a JSONL file (`--input-file`, optional `--model` for content-only lines, `--window` `15m`/`1h`/`24h`) |
 | `sference batch stream` | Submit, wait, print **JSONL results on stdout** (see below) |
 | `sference batch status` | Get one batch (`--batch-id`, `--json`) |
 | `sference batch wait` | Poll until terminal state (`--batch-id`, `--poll-interval`, `--timeout`, `--json`) |
@@ -145,7 +143,7 @@ Long-lived **streams** are separate from **batches**: you create a stream, submi
 
 | Command | Description |
 |---------|-------------|
-| `sference stream create` | Create a stream (`--name`, `--window` `1h` or `24h`, `--json`) |
+| `sference stream create` | Create a stream (`--name`, `--window` `15m`/`1h`/`24h`, `--json`) |
 | `sference stream list` | List streams (`--json`) |
 | `sference stream status` | Full detail + counters (`--stream-id`, `--json`) |
 | `sference stream submit` | Create responses from JSONL via `POST /v1/responses` per line (`metadata.stream_id` set automatically; `--stream-id`, `--input-file`, `--model` required for content-only lines) — per line: OpenAI batch-style `{custom_id?, method, url, body}` or content-only `{content}` |
@@ -223,9 +221,19 @@ sference batch stream --input-file fixtures/example_batch.jsonl --poll-interval
 
 The SDK and CLI accept two line shapes (see also [`fixtures/example_batch.jsonl`](fixtures/example_batch.jsonl)):
 
-1. **OpenAI-compatible:** each line has `custom_id`, `method`, `url`, and `body` (e.g. chat completions payload with per-line `model`). The CLI `--model` flag is ignored for these lines (a warning may be emitted by the SDK).
+1. **OpenAI-compatible envelope:** each line has `custom_id`, `method`, `url`, and `body`. The API receives only `custom_id` + inner `body` (`method`/`url` are ignored). Inner `body` may use:
+   - **Chat completions:** `messages` (non-empty array), plus optional `temperature`, `max_tokens`, `tools`, …
+   - **Responses API:** `input` (string or message array), optional `instructions`, `max_output_tokens`, … — normalized to chat format at batch create.
 2. **Content-only:** each line is `{"content": "..."}`. Then **`--model` is required** on submit/stream.
 
+Invalid row bodies return **HTTP 400** at create with `requests[i]` and `custom_id` in the error message (nothing is enqueued). Do not set `background: true` inside batch row bodies.
+
+Example Responses-shaped JSONL line:
+
+```json
+{"custom_id":"r2","method":"POST","url":"/v1/responses","body":{"model":"Qwen/Qwen3.6-35B-A3B","input":[{"role":"user","content":"hi"}],"max_output_tokens":256}}
+```
+
 ---
 
 ## Python SDK

{sference_cli-0.1.1 → sference_cli-0.1.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "sference-cli"
-version = "0.1.1"
+version = "0.1.2"
 description = "sference command-line interface"
 readme = "README.md"
 requires-python = ">=3.10"
@@ -23,7 +23,7 @@ classifiers = [
 ]
 dependencies = [
     "typer>=0.24.1",
-    "sference-sdk>=0.1.1",
+    "sference-sdk>=0.1.2",
 ]
 
 [project.urls]

{sference_cli-0.1.1 → sference_cli-0.1.2}/sference_cli/launch.py

@@ -12,7 +12,7 @@ from typing import Optional
 
 import typer
 
-DEFAULT_LAUNCH_MODEL = "moonshotai/Kimi-K2.6"
+DEFAULT_LAUNCH_MODEL = "moonshotai/Kimi-K2.7-Code"
 DEFAULT_API_BASE_URL = "https://api.sference.com"
 
 
@@ -301,5 +301,4 @@ def register_launch_commands(app: typer.Typer) -> None:
             dry_run=dry_run,
         )
 
-    # Hidden from `sference --help`; documented for Claude Code / Pi setup only.
-    app.add_typer(launch_app, name="launch", hidden=True)
+    app.add_typer(launch_app, name="launch")

{sference_cli-0.1.1 → sference_cli-0.1.2}/sference_cli/main.py

@@ -13,7 +13,7 @@ from typing import Any, Optional, TypeVar
 import typer
 
 import sference_sdk
-from sference_sdk import ApiError, SferenceClient
+from sference_sdk import COMPLETION_WINDOWS, ApiError, SferenceClient
 
 from sference_cli import __version__ as _CLI_VERSION
 from sference_sdk.checkpoint import clear_checkpoint, load_checkpoint, save_checkpoint
@@ -151,15 +151,10 @@ def _list_models(*, client: SferenceClient, as_json: bool) -> None:
     typer.echo(json.dumps(payload, indent=None if as_json else 2))
 
 
-def _mvp_batch_window_only(value: str) -> str:
-    if value != "24h":
-        raise typer.BadParameter('Only "24h" is supported in MVP.', param_hint="--window")
-    return value
-
-
-def _stream_window_only(value: str) -> str:
-    if value not in ("1h", "24h"):
-        raise typer.BadParameter('Window must be "1h" or "24h".', param_hint="--window")
+def _window_choice(value: str) -> str:
+    if value not in COMPLETION_WINDOWS:
+        allowed = ", ".join(f'"{w}"' for w in COMPLETION_WINDOWS)
+        raise typer.BadParameter(f"Window must be one of {allowed}.", param_hint="--window")
     return value
 
 
@@ -506,8 +501,8 @@ def batch_submit(
     window: str = typer.Option(
         "24h",
         "--window",
-        help='Batch SLA window (MVP: only "24h").',
-        callback=_mvp_batch_window_only,
+        help='Batch SLA window: "15m", "1h", or "24h".',
+        callback=_window_choice,
     ),
     base_url: str = typer.Option("https://api.sference.com"),
     as_json: bool = typer.Option(False, "--json"),
@@ -537,8 +532,8 @@ def batch_stream(
     window: str = typer.Option(
         "24h",
         "--window",
-        help='Batch SLA window (MVP: only "24h").',
-        callback=_mvp_batch_window_only,
+        help='Batch SLA window: "15m", "1h", or "24h".',
+        callback=_window_choice,
     ),
     poll_interval: float = typer.Option(2.0, "--poll-interval", help="Seconds between status polls while the batch is running."),
     no_cache: bool = typer.Option(
@@ -653,8 +648,8 @@ def stream_create(
     window: str = typer.Option(
         "24h",
         "--window",
-        help='Per-item SLA window: "1h" or "24h" (recorded only in MVP).',
-        callback=_stream_window_only,
+        help='Per-item SLA window: "15m", "1h", or "24h" (default for items in this stream).',
+        callback=_window_choice,
     ),
     base_url: str = typer.Option("https://api.sference.com"),
     as_json: bool = typer.Option(False, "--json"),