rockycode 0.1.1__tar.gz → 0.1.2__tar.gz

rockycode-0.1.2/.github/workflows/release.yml

@@ -0,0 +1,27 @@
+name: Release
+
+# One job: build sdist+wheel on ubuntu, publish to PyPI via Trusted Publishing.
+# The old macos-13 matrix (built a cbor2 x86_64-mac wheel for the release page)
+# is gone: the runner label was retired, so it never actually ran — 0.1.0 and
+# 0.1.1 both sat in its queue until cancelled and shipped fine without it — and
+# cbor2 still publishes no x86_64-mac wheel to attach anyway (arm64 only; an
+# Intel-Mac install builds it from source via Rust, PyPI-side, as it always
+# has). GitHub releases are created by hand: gh release create --notes-file.
+on:
+  push:
+    tags: ["v*"]
+  workflow_dispatch:
+
+jobs:
+  pypi:
+    runs-on: ubuntu-latest
+    permissions:
+      id-token: write            # OIDC — this is what Trusted Publishing uses
+    steps:
+      - uses: actions/checkout@v4
+      - name: Setup uv
+        uses: astral-sh/setup-uv@v5
+      - name: Build rockycode (sdist + wheel — only our project lands in dist/)
+        run: uv build
+      - name: Publish to PyPI (Trusted Publishing, no token)
+        uses: pypa/gh-action-pypi-publish@release/v1

{rockycode-0.1.1 → rockycode-0.1.2}/CHANGELOG.md

@@ -9,6 +9,25 @@ change between minor versions.
 
 _Nothing yet._
 
+## [0.1.2] — `--version` flag, GA price tables
+
+### Added
+- `rockycode --version` / `-V` prints the installed version. One version
+  source: package metadata (pyproject) — the serve handshake reports the same
+  value instead of a hardcoded string.
+
+### Changed
+- DeepSeek price tables refreshed to the GA snapshots (V4-Flash-0731 /
+  V4-Pro-0813), both USD and CNY, verified 2026-08-20 at the source;
+  peak-valley billing confirmed live (2× in the published UTC windows).
+- README results updated: `deepseek-v4-flash` GA carries three clean full-500
+  SWE-bench Verified rounds (88.8% average, 95.4% pass@3); the
+  `deepseek-v4-pro` column is explicitly marked preview (pre-0813).
+
+### CI
+- Release workflow reduced to the single ubuntu PyPI Trusted-Publishing job;
+  the dead macos-13 matrix (retired runner, never ran) is gone.
+
 ## [0.1.1] — session artifacts, images in chat, real sandbox cancel
 
 ### Added

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

@@ -1,6 +1,6 @@
-Metadata-Version: 2.4
+Metadata-Version: 2.5
 Name: rockycode
-Version: 0.1.1
+Version: 0.1.2
 Summary: A coding agent harness, benchmarked on SWE-bench Verified. amaze!
 Author: rockycode contributors
 License: MIT
@@ -42,8 +42,8 @@ Built for the DeepSeek V4 series, with a unique research mode, bench-tested, and
 
 [English](README.md) · [简体中文](README.zh-CN.md)
 
-![SWE-bench Verified](https://img.shields.io/badge/SWE--bench_Verified-79.8%25_V4--flash-7d5cc6)
-![V4-pro preview](https://img.shields.io/badge/V4--pro_preview-81.8%25_pass@3-8d6cd0)
+![SWE-bench Verified](https://img.shields.io/badge/SWE--bench_Verified-88.8%25_V4--flash_GA-7d5cc6)
+![V4-pro preview](https://img.shields.io/badge/V4--pro_preview-74.9%25-8d6cd0)
 ![Python](https://img.shields.io/badge/Python-3.11%2B-9d7cd8)
 ![License](https://img.shields.io/badge/License-MIT-a9b1d6)
 
@@ -77,31 +77,37 @@ against it.
 
 ## Results — SWE-bench Verified
 
-Full-set numbers: **independent full-500 runs** (three each for
-`deepseek-v4-pro` and `minimax-m3`, one so far for `deepseek-v4-flash`), same
+Full-set numbers: **independent full-500 runs**, three rounds per model, same
 harness and config for all (100-step cap, 32,768 max output tokens, reasoning
 effort `max`, thinking on), scored with the official SWE-bench harness. No
-tuning against the tasks.
+tuning against the tasks. Mind the versions: the `deepseek-v4-flash` column is
+the **GA release** (V4-Flash-0731), while the `deepseek-v4-pro` rounds ran
+before the GA V4-Pro-0813 shipped — that column is the **preview** pro, so the
+flash/pro gap reflects a version difference, not a same-vintage comparison.
 
-| run | `deepseek-v4-pro` | `deepseek-v4-flash` | `minimax-m3` |
+| run | `deepseek-v4-flash` (GA) | `deepseek-v4-pro` (preview) | `minimax-m3` |
 |---|---|---|---|
-| round 1 | 75.6% (378/500) | 79.8% (399/500) | 72.8% (364/500) |
-| round 2 | 74.8% (374/500) | — | 71.2% (356/500) |
-| round 3 | 74.4% (372/500) | — | 70.0% (350/500) |
-| **average** | **74.9%** | **79.8%** *(1 run)* | **71.3%** |
-| union of runs (pass@3) | 81.8% (409/500) | — | 83.6% (418/500) |
+| round 1 | 90.0% (450/500) | 75.6% (378/500) | 72.8% (364/500) |
+| round 2 | 88.6% (443/500) | 74.8% (374/500) | 71.2% (356/500) |
+| round 3 | 87.8% (439/500) | 74.4% (372/500) | 70.0% (350/500) |
+| **average** | **88.8%** | **74.9%** | **71.3%** |
+| union of runs (pass@3) | 95.4% (477/500) | 81.8% (409/500) | 83.6% (418/500) |
 
 Read the two summary rows differently. The **average** is the
 leaderboard-comparable number — each round is an independent single-pass run
 over the full 500. The **union** is pass@3: tasks solved by at least one
-round. The gap between them (~7 points for DeepSeek, ~12 for MiniMax) is
-run-to-run variance, not capability — the models already reach these tasks
-under this harness, they just don't hold them every run. Closing that gap
-(verify-before-finish gating and run selection, not more prompting) is the
-current line of work. For reference, DeepSeek reports 80.6% with its own
-scaffold; the three-run union crosses that mark. Per-round breakdowns:
+round. The gap between them (~7 points for both DeepSeek models, ~12 for
+MiniMax) is run-to-run variance, not capability — the models already reach
+these tasks under this harness, they just don't hold them every run. Closing
+that gap (verify-before-finish gating and run selection, not more prompting)
+is the current line of work. For reference, DeepSeek reported 80.6% for V4
+Pro (Preview) with its own scaffold. Per-round breakdowns:
 [@rockycode_ai](https://x.com/rockycode_ai).
 
+Two earlier flash rounds (79.8% — previously listed here — and 81.2%) hit
+local network outages mid-run, visible as contiguous blocks of empty-patch
+tasks; they were replaced by clean re-runs rather than averaged in.
+
 We plan to add **DeepSWE-bench** support as well — currently in progress.
 
 ## Getting started
@@ -121,6 +127,10 @@ uv tool install rockycode     # recommended — puts the `rockycode` command on
 rockycode                     # the first run walks you through API-key setup
 ```
 
+Already installed and a new release is out? **`uv tool upgrade rockycode`** —
+note that re-running `uv tool install` does NOT upgrade: it sees the existing
+install and quietly keeps the old version.
+
 Don't have uv yet? One command installs it:
 `curl -LsSf https://astral.sh/uv/install.sh | sh` — Windows and other options
 in the [uv install docs](https://docs.astral.sh/uv/getting-started/installation/).

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

@@ -9,8 +9,8 @@ Built for the DeepSeek V4 series, with a unique research mode, bench-tested, and
 
 [English](README.md) · [简体中文](README.zh-CN.md)
 
-![SWE-bench Verified](https://img.shields.io/badge/SWE--bench_Verified-79.8%25_V4--flash-7d5cc6)
-![V4-pro preview](https://img.shields.io/badge/V4--pro_preview-81.8%25_pass@3-8d6cd0)
+![SWE-bench Verified](https://img.shields.io/badge/SWE--bench_Verified-88.8%25_V4--flash_GA-7d5cc6)
+![V4-pro preview](https://img.shields.io/badge/V4--pro_preview-74.9%25-8d6cd0)
 ![Python](https://img.shields.io/badge/Python-3.11%2B-9d7cd8)
 ![License](https://img.shields.io/badge/License-MIT-a9b1d6)
 
@@ -44,31 +44,37 @@ against it.
 
 ## Results — SWE-bench Verified
 
-Full-set numbers: **independent full-500 runs** (three each for
-`deepseek-v4-pro` and `minimax-m3`, one so far for `deepseek-v4-flash`), same
+Full-set numbers: **independent full-500 runs**, three rounds per model, same
 harness and config for all (100-step cap, 32,768 max output tokens, reasoning
 effort `max`, thinking on), scored with the official SWE-bench harness. No
-tuning against the tasks.
+tuning against the tasks. Mind the versions: the `deepseek-v4-flash` column is
+the **GA release** (V4-Flash-0731), while the `deepseek-v4-pro` rounds ran
+before the GA V4-Pro-0813 shipped — that column is the **preview** pro, so the
+flash/pro gap reflects a version difference, not a same-vintage comparison.
 
-| run | `deepseek-v4-pro` | `deepseek-v4-flash` | `minimax-m3` |
+| run | `deepseek-v4-flash` (GA) | `deepseek-v4-pro` (preview) | `minimax-m3` |
 |---|---|---|---|
-| round 1 | 75.6% (378/500) | 79.8% (399/500) | 72.8% (364/500) |
-| round 2 | 74.8% (374/500) | — | 71.2% (356/500) |
-| round 3 | 74.4% (372/500) | — | 70.0% (350/500) |
-| **average** | **74.9%** | **79.8%** *(1 run)* | **71.3%** |
-| union of runs (pass@3) | 81.8% (409/500) | — | 83.6% (418/500) |
+| round 1 | 90.0% (450/500) | 75.6% (378/500) | 72.8% (364/500) |
+| round 2 | 88.6% (443/500) | 74.8% (374/500) | 71.2% (356/500) |
+| round 3 | 87.8% (439/500) | 74.4% (372/500) | 70.0% (350/500) |
+| **average** | **88.8%** | **74.9%** | **71.3%** |
+| union of runs (pass@3) | 95.4% (477/500) | 81.8% (409/500) | 83.6% (418/500) |
 
 Read the two summary rows differently. The **average** is the
 leaderboard-comparable number — each round is an independent single-pass run
 over the full 500. The **union** is pass@3: tasks solved by at least one
-round. The gap between them (~7 points for DeepSeek, ~12 for MiniMax) is
-run-to-run variance, not capability — the models already reach these tasks
-under this harness, they just don't hold them every run. Closing that gap
-(verify-before-finish gating and run selection, not more prompting) is the
-current line of work. For reference, DeepSeek reports 80.6% with its own
-scaffold; the three-run union crosses that mark. Per-round breakdowns:
+round. The gap between them (~7 points for both DeepSeek models, ~12 for
+MiniMax) is run-to-run variance, not capability — the models already reach
+these tasks under this harness, they just don't hold them every run. Closing
+that gap (verify-before-finish gating and run selection, not more prompting)
+is the current line of work. For reference, DeepSeek reported 80.6% for V4
+Pro (Preview) with its own scaffold. Per-round breakdowns:
 [@rockycode_ai](https://x.com/rockycode_ai).
 
+Two earlier flash rounds (79.8% — previously listed here — and 81.2%) hit
+local network outages mid-run, visible as contiguous blocks of empty-patch
+tasks; they were replaced by clean re-runs rather than averaged in.
+
 We plan to add **DeepSWE-bench** support as well — currently in progress.
 
 ## Getting started
@@ -88,6 +94,10 @@ uv tool install rockycode     # recommended — puts the `rockycode` command on
 rockycode                     # the first run walks you through API-key setup
 ```
 
+Already installed and a new release is out? **`uv tool upgrade rockycode`** —
+note that re-running `uv tool install` does NOT upgrade: it sees the existing
+install and quietly keeps the old version.
+
 Don't have uv yet? One command installs it:
 `curl -LsSf https://astral.sh/uv/install.sh | sh` — Windows and other options
 in the [uv install docs](https://docs.astral.sh/uv/getting-started/installation/).

{rockycode-0.1.1 → rockycode-0.1.2}/README.zh-CN.md

@@ -9,8 +9,8 @@
 
 [English](README.md) · [简体中文](README.zh-CN.md)
 
-![SWE-bench Verified](https://img.shields.io/badge/SWE--bench_Verified-79.8%25_V4--flash-7d5cc6)
-![V4-pro preview](https://img.shields.io/badge/V4--pro_preview-81.8%25_pass@3-8d6cd0)
+![SWE-bench Verified](https://img.shields.io/badge/SWE--bench_Verified-88.8%25_V4--flash_GA-7d5cc6)
+![V4-pro preview](https://img.shields.io/badge/V4--pro_preview-74.9%25-8d6cd0)
 ![Python](https://img.shields.io/badge/Python-3.11%2B-9d7cd8)
 ![License](https://img.shields.io/badge/License-MIT-a9b1d6)
 
@@ -37,17 +37,19 @@ rockycode 是一个编程智能体 harness，为 DeepSeek V4 系列适配，也
 
 ## 能力量化：SWE bench
 
-完整 500 题的成绩：**独立的完整 500 题运行**（`deepseek-v4-pro` 与 `minimax-m3` 各 3 轮，`deepseek-v4-flash` 目前 1 轮），所有模型使用完全相同的 harness 与配置（步数上限 100、单次最大输出 32,768 token、推理力度 `max`、thinking 开启），由官方 SWE-bench harness 打分，未针对任务做任何调优。
+完整 500 题的成绩：**独立的完整 500 题运行**，每个模型各 3 轮，所有模型使用完全相同的 harness 与配置（步数上限 100、单次最大输出 32,768 token、推理力度 `max`、thinking 开启），由官方 SWE-bench harness 打分，未针对任务做任何调优。注意版本差异：`deepseek-v4-flash` 一列是**正式版（GA，V4-Flash-0731）**；而 `deepseek-v4-pro` 的三轮在 GA 版 V4-Pro-0813 发布之前运行，该列是 **preview 版**成绩 —— flash 与 pro 的差距里含有版本差，不是同代对比。
 
-| 轮次 | `deepseek-v4-pro` | `deepseek-v4-flash` | `minimax-m3` |
+| 轮次 | `deepseek-v4-flash`（GA） | `deepseek-v4-pro`（preview） | `minimax-m3` |
 |---|---|---|---|
-| 第 1 轮 | 75.6%（378/500） | 79.8%（399/500） | 72.8%（364/500） |
-| 第 2 轮 | 74.8%（374/500） | — | 71.2%（356/500） |
-| 第 3 轮 | 74.4%（372/500） | — | 70.0%（350/500） |
-| **平均** | **74.9%** | **79.8%**（1 轮） | **71.3%** |
-| 多轮并集（pass@3） | 81.8%（409/500） | — | 83.6%（418/500） |
+| 第 1 轮 | 90.0%（450/500） | 75.6%（378/500） | 72.8%（364/500） |
+| 第 2 轮 | 88.6%（443/500） | 74.8%（374/500） | 71.2%（356/500） |
+| 第 3 轮 | 87.8%（439/500） | 74.4%（372/500） | 70.0%（350/500） |
+| **平均** | **88.8%** | **74.9%** | **71.3%** |
+| 多轮并集（pass@3） | 95.4%（477/500） | 81.8%（409/500） | 83.6%（418/500） |
 
-两行汇总要分开读。**平均值**是可以与排行榜对比的数字 —— 每一轮都是独立的单次完整 500 题。**并集**是 pass@3：至少被某一轮解出的任务。两者之间的差距（DeepSeek 约 7 个点、MiniMax 约 12 个点）是轮次间方差，不是能力上限 —— 模型在这套 harness 下已经"够得着"这些任务，只是无法每一轮都稳住。当前的工作重心就是收掉这个差距：finish 前的验证门控与多轮选择，而不是继续改提示词。作为参照：DeepSeek 用自家 scaffold 报告 80.6%，三轮并集已越过这个数字。逐轮拆解请关注我们的X账号（@rockycode_ai）。
+两行汇总要分开读。**平均值**是可以与排行榜对比的数字 —— 每一轮都是独立的单次完整 500 题。**并集**是 pass@3：至少被某一轮解出的任务。两者之间的差距（DeepSeek 两个模型各约 7 个点、MiniMax 约 12 个点）是轮次间方差，不是能力上限 —— 模型在这套 harness 下已经"够得着"这些任务，只是无法每一轮都稳住。当前的工作重心就是收掉这个差距：finish 前的验证门控与多轮选择，而不是继续改提示词。作为参照：DeepSeek 用自家 scaffold 报告 V4 Pro (Preview) 为 80.6%。逐轮拆解请关注我们的X账号（@rockycode_ai）。
+
+另有两轮较早的 flash-preview 运行（79.8% —— 即此前列在这里的那一轮 —— 与 81.2%）在运行途中遭遇本地网络中断，表现为连续任务块返回空补丁；这两轮已用干净的重跑替换，未计入平均。
 
 我们计划支持DeepSWE bench，目前还在调试中。
 
@@ -62,6 +64,9 @@ uv tool install rockycode     # 推荐 —— `rockycode` 命令直接进 PATH
 rockycode                     # 首次运行会引导你完成 API key 设置
 ```
 
+已经装过、想升到新版本？用 **`uv tool upgrade rockycode`** ——
+注意重复执行 `uv tool install` 不会升级：它看到已有安装就静默保留旧版本。
+
 还没有 uv？一条命令安装：
 `curl -LsSf https://astral.sh/uv/install.sh | sh`（Windows 及其他方式见
 [uv 安装文档](https://docs.astral.sh/uv/getting-started/installation/)）。

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

@@ -1,6 +1,6 @@
 [project]
 name = "rockycode"
-version = "0.1.1"
+version = "0.1.2"
 description = "A coding agent harness, benchmarked on SWE-bench Verified. amaze!"
 requires-python = ">=3.11"
 readme = "README.md"

rockycode-0.1.2/rockycode/__init__.py

@@ -0,0 +1,11 @@
+# Version comes from installed package metadata (single source: pyproject) —
+# a hardcoded string here shipped stale once ("0.1.0" inside the 0.1.1 wheel).
+# Lazy module __getattr__ keeps `import rockycode` free of the metadata scan.
+def __getattr__(name: str):
+    if name == "__version__":
+        from importlib.metadata import PackageNotFoundError, version
+        try:
+            return version("rockycode")
+        except PackageNotFoundError:  # running from a bare checkout, uninstalled
+            return "unknown"
+    raise AttributeError(name)

{rockycode-0.1.1 → rockycode-0.1.2}/rockycode/cli.py

@@ -29,7 +29,7 @@ bootstrap_credentials()
 # `rockycode` without args → `rockycode chat`; a leading flag (`rockycode
 # --resume …`, `rockycode --yolo`) reaches chat too — nobody should have to
 # remember to type "chat" first.
-_TOP_LEVEL_FLAGS = {"--help", "-h", "--version", "--install-completion", "--show-completion"}
+_TOP_LEVEL_FLAGS = {"--help", "-h", "--version", "-V", "--install-completion", "--show-completion"}
 if len(sys.argv) == 1:
     sys.argv.append("chat")
 elif sys.argv[1].startswith("-") and sys.argv[1] not in _TOP_LEVEL_FLAGS:
@@ -67,8 +67,19 @@ def _env_int(name: str, default: int) -> int:
         return default
 
 
+def _version_flag(value: bool) -> None:
+    if value:
+        import rockycode
+        print(f"rockycode {rockycode.__version__}")  # plain print: script-friendly
+        raise typer.Exit()
+
+
 @app.callback()
-def _root() -> None:
+def _root(
+    version: bool = typer.Option(
+        False, "--version", "-V", callback=_version_flag, is_eager=True,
+        help="Print the version and exit."),
+) -> None:
     """Force Typer into multi-command mode."""
 
 

{rockycode-0.1.1 → rockycode-0.1.2}/rockycode/engine/server.py

@@ -413,9 +413,10 @@ async def run_server(
                 # chain: env, .env files serve loaded, keychain) so the client's
                 # "setup required" card reflects reality instead of guessing
                 # from the editor's own environment.
+                import rockycode
                 from rockycode.onboarding import is_configured
                 _write_line(_response(msg_id, {
-                    "version": "0.1.1",
+                    "version": rockycode.__version__,
                     "session_id": sess.session_id,
                     "model": model,
                     "configured": is_configured(),

{rockycode-0.1.1 → rockycode-0.1.2}/rockycode/pricing.py

@@ -9,7 +9,9 @@ DeepSeek changes them:
     https://api-docs.deepseek.com/quick_start/pricing        (USD table)
     https://api-docs.deepseek.com/zh-cn/quick_start/pricing   (CNY table)
 
-Verified: 2026-07-04 (per 1,000,000 tokens).
+Verified: 2026-08-20 (per 1,000,000 tokens). Tables list the GA snapshots
+DeepSeek-V4-Flash-0731 and DeepSeek-V4-Pro-0813; both are billed off-peak base
+with a 2x peak surcharge.
 
 To update WITHOUT editing the install (survives upgrades), drop a
 ~/.rockycode/pricing.toml that overrides any values below. It is a FILE on
@@ -24,18 +26,17 @@ from datetime import datetime, time as dtime, timezone
 from pathlib import Path
 
 PRICING_SOURCE_URL = "https://api-docs.deepseek.com/quick_start/pricing"
-PRICING_VERIFIED = "2026-07-04"
+PRICING_VERIFIED = "2026-08-20"
 OVERRIDE_PATH = Path.home() / ".rockycode" / "pricing.toml"
 
 # Official list prices, per 1M tokens, each currency from its OWN table.
 DEFAULT_PRICING: dict = {
     "peak": {
-        # DeepSeek peak-valley pricing (announced): peak-hour rate = 2x regular,
-        # for ALL billing items, during these UTC windows — 01:00–04:00 and
-        # 06:00–10:00. It starts MID-JULY 2026, so the logic is wired but gated
-        # behind effective_date: before that date nothing is surcharged even
-        # inside a window, and it auto-activates on the date with no code change.
-        # Confirm the exact start date and adjust here or in the override file.
+        # DeepSeek peak-valley pricing, LIVE and confirmed at the source
+        # 2026-08-20: peak-hour rate = 2x the off-peak base below, for ALL
+        # billing items, during these UTC windows — 01:00–04:00 and 06:00–10:00
+        # (Beijing 9:00–12:00 / 14:00–18:00). effective_date stays so usage
+        # from before the mid-July start keeps pricing at base.
         "enabled": True,
         "effective_date": "2026-07-15",  # UTC; peak surcharge does not apply before this
         "multiplier": 2.0,
@@ -45,13 +46,14 @@ DEFAULT_PRICING: dict = {
         ],
     },
     "models": {
-        "deepseek-v4-pro": {
-            "usd": {"in_hit": 0.003625, "in_miss": 0.435, "out": 0.87},
-            "cny": {"in_hit": 0.025, "in_miss": 3.0, "out": 6.0},
+        # Off-peak base rates; peak windows above bill at multiplier x base.
+        "deepseek-v4-pro": {  # GA snapshot V4-Pro-0813
+            "usd": {"in_hit": 0.022, "in_miss": 0.66, "out": 1.98},
+            "cny": {"in_hit": 0.15, "in_miss": 4.5, "out": 13.5},
         },
-        "deepseek-v4-flash": {
-            "usd": {"in_hit": 0.0028, "in_miss": 0.14, "out": 0.28},
-            "cny": {"in_hit": 0.02, "in_miss": 1.0, "out": 2.0},
+        "deepseek-v4-flash": {  # GA snapshot V4-Flash-0731
+            "usd": {"in_hit": 0.007, "in_miss": 0.22, "out": 0.66},
+            "cny": {"in_hit": 0.05, "in_miss": 1.5, "out": 4.5},
         },
     },
     "fallback_model": "deepseek-v4-pro",  # unknown model → price as pro (conservative)

{rockycode-0.1.1 → rockycode-0.1.2}/rockycode-vscode/package.json

@@ -2,7 +2,7 @@
   "name": "rockycode",
   "displayName": "rockycode",
   "description": "AI coding agent — chat, inline tools, diff previews, and in-panel Enter/arrow tool approvals. Powered by the rockycode CLI (DeepSeek).",
-  "version": "0.1.1",
+  "version": "0.1.2",
   "publisher": "rockycode",
   "license": "MIT",
   "icon": "media/icon.png",

{rockycode-0.1.1 → rockycode-0.1.2}/tests/smoke_pricing.py

@@ -18,10 +18,10 @@ def test_dual_currency():
     led = UsageLedger()  # peak disabled by default
     led.add("deepseek-v4-pro", _1M, at=OFFPEAK)
     led.add("deepseek-v4-flash", _1M, at=OFFPEAK)
-    # USD from the USD table: pro 0.435+0.87, flash 0.14+0.28
-    assert abs(led.cost("usd") - (0.435 + 0.87 + 0.14 + 0.28)) < 1e-9, led.cost("usd")
-    # CNY from the CNY table — independent numbers, NOT usd*ratio: pro 3+6, flash 1+2
-    assert abs(led.cost("cny") - (3.0 + 6.0 + 1.0 + 2.0)) < 1e-9, led.cost("cny")
+    # USD from the USD table: pro 0.66+1.98, flash 0.22+0.66
+    assert abs(led.cost("usd") - (0.66 + 1.98 + 0.22 + 0.66)) < 1e-9, led.cost("usd")
+    # CNY from the CNY table — independent numbers, NOT usd*ratio: pro 4.5+13.5, flash 1.5+4.5
+    assert abs(led.cost("cny") - (4.5 + 13.5 + 1.5 + 4.5)) < 1e-9, led.cost("cny")
     assert led.configured("usd") and led.configured("cny")
     print(f"dual-currency: ${led.cost('usd'):.3f} / ¥{led.cost('cny'):.1f} (independent tables)  ok")
 
@@ -29,13 +29,13 @@ def test_dual_currency():
 def test_peak():
     led = UsageLedger()  # ships peak enabled (2x, windows 01–04 & 06–10 UTC, from mid-July)
     led.add("deepseek-v4-flash", _1M, at=PEAK)
-    assert abs(led.cost("usd") - (0.14 + 0.28) * 2) < 1e-9, led.cost("usd")   # 2x in-window
+    assert abs(led.cost("usd") - (0.22 + 0.66) * 2) < 1e-9, led.cost("usd")   # 2x in-window
     off = UsageLedger()
     off.add("deepseek-v4-flash", _1M, at=OFFPEAK)
-    assert abs(off.cost("usd") - (0.14 + 0.28)) < 1e-9, off.cost("usd")       # off-peak base
+    assert abs(off.cost("usd") - (0.22 + 0.66)) < 1e-9, off.cost("usd")       # off-peak base
     pre = UsageLedger()
     pre.add("deepseek-v4-flash", _1M, at=PRE_EFFECT)
-    assert abs(pre.cost("usd") - (0.14 + 0.28)) < 1e-9, pre.cost("usd")       # in-window but before mid-July
+    assert abs(pre.cost("usd") - (0.22 + 0.66)) < 1e-9, pre.cost("usd")       # in-window but before mid-July
     print("peak-hour: 2x in-window after the mid-July start; base before it and off-peak  ok")
 
 
@@ -45,7 +45,7 @@ def test_override():
     p.write_text("[models.deepseek-v4-flash.usd]\nout = 9.99\n")
     pricing = load_pricing(override_path=p)
     assert pricing["models"]["deepseek-v4-flash"]["usd"]["out"] == 9.99      # overridden
-    assert pricing["models"]["deepseek-v4-flash"]["usd"]["in_miss"] == 0.14  # untouched keeps default
+    assert pricing["models"]["deepseek-v4-flash"]["usd"]["in_miss"] == 0.22  # untouched keeps default
     print("override: ~/.rockycode/pricing.toml merges over the built-in table  ok")
 
 
@@ -54,7 +54,7 @@ def test_cache_cheap():
     led.add("deepseek-v4-pro",
             {"prompt_tokens": 1_000_000, "prompt_cache_hit_tokens": 1_000_000, "completion_tokens": 0},
             at=OFFPEAK)
-    assert led.cost("usd") < 0.02, led.cost("usd")
+    assert led.cost("usd") < 0.03, led.cost("usd")  # 1M hit tokens at $0.022 vs $0.66 as misses
     print("cache hits cost far less than misses  ok")
 
 

{rockycode-0.1.1 → rockycode-0.1.2}/tests/smoke_serve.py

@@ -66,7 +66,7 @@ def test_serve_jsonrpc_handshake():
             proc.stdin.write(_request("initialize"))
             proc.stdin.flush()
             resp = _read_line(proc)
-            assert resp.get("result", {}).get("version") == "0.1.1", \
+            assert resp.get("result", {}).get("version") == "0.1.2", \
                 f"unexpected init response: {resp}"
             sid = resp.get("result", {}).get("session_id")
             assert sid
@@ -107,7 +107,7 @@ def test_serve_survives_malformed_messages():
             proc.stdin.write(_request("initialize"))
             proc.stdin.flush()
             resp = _read_line(proc)
-            assert resp.get("result", {}).get("version") == "0.1.1", \
+            assert resp.get("result", {}).get("version") == "0.1.2", \
                 f"server did not survive malformed input: {resp}"
             proc.stdin.write(_request("shutdown", msg_id=2))
             proc.stdin.flush()
@@ -175,6 +175,17 @@ def test_event_to_notification():
         assert n["params"]["session_id"] == sid
 
 
+def test_cli_version_flag():
+    """`rockycode --version` prints the SAME version the serve handshake
+    reports — both derive from package metadata, and this literal is part of
+    the release bump ritual (matches the handshake asserts above)."""
+    from typer.testing import CliRunner
+
+    from rockycode.cli import app
+    result = CliRunner().invoke(app, ["--version"])
+    assert result.exit_code == 0 and "rockycode 0.1.2" in result.output, result.output
+
+
 if __name__ == "__main__":
     test_serve_jsonrpc_handshake()
     print("PASS test_serve_jsonrpc_handshake")
@@ -188,4 +199,7 @@ if __name__ == "__main__":
     test_event_to_notification()
     print("PASS test_event_to_notification")
 
+    test_cli_version_flag()
+    print("PASS test_cli_version_flag")
+
     print("\nOK — all serve smoke tests passed")

{rockycode-0.1.1 → rockycode-0.1.2}/uv.lock

@@ -2239,7 +2239,7 @@ wheels = [
 
 [[package]]
 name = "rockycode"
-version = "0.1.1"
+version = "0.1.2"
 source = { editable = "." }
 dependencies = [
     { name = "aiohttp" },

rockycode-0.1.1/.github/workflows/release.yml

@@ -1,73 +0,0 @@
-name: Release
-
-on:
-  push:
-    tags: ["v*"]
-  workflow_dispatch:
-
-jobs:
-  release:
-    runs-on: macos-13  # x86_64 Mac — the only platform missing cbor2 wheels
-    strategy:
-      matrix:
-        python: ["3.11", "3.12", "3.13"]
-    steps:
-      - uses: actions/checkout@v4
-
-      - name: Setup uv
-        uses: astral-sh/setup-uv@v5
-        with:
-          enable-cache: true
-          python-version: ${{ matrix.python }}
-
-      - name: Install Rust (cbor2 needs it on macOS x86_64)
-        run: rustup update stable
-
-      - name: Build rockycode wheel (only once)
-        if: matrix.python == '3.11'
-        run: uv build
-
-      - name: Build cbor2 wheel
-        run: uv pip wheel cbor2==6.1.2 --no-deps -w dist/
-
-      - name: Upload wheels
-        uses: actions/upload-artifact@v4
-        with:
-          name: wheels-${{ matrix.python }}
-          path: dist/
-
-  publish:
-    needs: release
-    runs-on: ubuntu-latest
-    steps:
-      - name: Download all wheels
-        uses: actions/download-artifact@v4
-        with:
-          path: dist/
-          pattern: wheels-*
-          merge-multiple: true
-
-      - name: Show artifacts
-        run: ls -la dist/
-
-      - name: Release
-        uses: softprops/action-gh-release@v2
-        with:
-          files: dist/*
-          generate_release_notes: true
-
-  # rockycode is pure-Python, so PyPI publishing builds on Ubuntu and does NOT
-  # depend on the macOS `release` job (which only exists for the cbor2 wheel on
-  # the GitHub-Release side). This keeps publishing off the slow macOS queue.
-  pypi:
-    runs-on: ubuntu-latest
-    permissions:
-      id-token: write            # OIDC — this is what Trusted Publishing uses
-    steps:
-      - uses: actions/checkout@v4
-      - name: Setup uv
-        uses: astral-sh/setup-uv@v5
-      - name: Build rockycode (sdist + wheel — only our project lands in dist/)
-        run: uv build
-      - name: Publish to PyPI (Trusted Publishing, no token)
-        uses: pypa/gh-action-pypi-publish@release/v1

rockycode-0.1.1/rockycode/__init__.py

@@ -1 +0,0 @@
-__version__ = "0.1.0"