pythinker-code 0.25.0__tar.gz → 0.26.0__tar.gz

{pythinker_code-0.25.0 → pythinker_code-0.26.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pythinker-code
-Version: 0.25.0
+Version: 0.26.0
 Summary: Pythinker Code is your next CLI agent.
 Keywords: cli,agent,ai,coding-assistant,llm,claude,openai,terminal
 Author: Mohamed Elkholy
@@ -115,16 +115,15 @@ It speaks the [**Agent Client Protocol (ACP)**](https://github.com/agentclientpr
 
 ---
 
-## 🆕 What's New in 0.25.0
+## 🆕 What's New in 0.26.0
 
-- **`Fetch` re-checks every redirect hop against the SSRF guard.** Redirects were followed without re-validating the destination, so a public URL could redirect to a link-local address (e.g. a cloud metadata endpoint) and slip past the guard. Every hop is now re-validated, closing the public→link-local bypass.
-- **Web domain allowlist for `Fetch` and `Search`.** Set `web.allowed_domains` to restrict which hosts the web tools may reach — `Fetch` (every redirect hop included) and `Search` reject anything off the list. Leave it unset to keep web access unrestricted.
-- **Crash-consistent background tasks.** Task state is serialised under a cross-process lock, terminal updates route through a single finalizer, and recovery reconciles records left divergent by a crash or kill without clobbering a live agent. Bash output is capped (default 50 MiB), kills escalate SIGTERM→SIGKILL, and aged terminal tasks are pruned (default 7 days).
-- **Calmer TUI with live tool feedback.** Tool calls show a "preparing" row during approval, stream shell output as a running tail, and render a live Markdown preview as the model writes. The todo list no longer renders twice mid-turn, links open through a detached launcher, and the terminal is restored to a sane state on `SIGTERM`/`SIGQUIT` and at exit.
-- **Steadier agent loop.** The model is nudged when a turn ends on a bare statement of intent, steered away from blocking on one background task while siblings run, and a `SetTodoList` whose todos arrive as a JSON string is parsed transparently instead of failing validation.
-- **Unified report rendering.** Code review, verify, and security-review output share one standardized renderer, including `report` blocks emitted by skills and agents.
+- **Install and update no longer 404 on a freshly published release.** Releases are now published as a prerelease and promoted to "latest" only once every platform asset has uploaded, and the curl/PowerShell installers wait for their archive + checksum before downloading — so `irm | iex` / `curl | bash` run mid-publish no longer fail on a missing installer.
+- **OpenCode Go model catalog refreshes on startup without a re-login.** Bundled catalog changes (new models, corrected provider shapes) now reach existing installs through the every-startup refresh, across both OpenAI- and Anthropic-shaped providers, while preserving your default model and thinking settings.
+- **Homebrew install shows the logo and fails fast on a missing tap token.** `brew install` now prints the robot-head banner via formula caveats, and the tap workflow reports an actionable error when `HOMEBREW_TAP_TOKEN` is missing instead of an opaque git failure.
+- **Markdown and report rendering hardened against real model output.** Code-span pipes inside tables no longer fracture the table, and a `report` block nested inside a documentation fence is no longer wrongly promoted — report fences are now extracted via a markdown-it AST walk.
+- **Steadier agent editing and a restored update prompt.** File-replace edit handling is hardened, subagent prompt persistence and the agent's working language are preserved, and the blocking pre-start update prompt is restored so it runs before every interactive session.
 
-Upgrade with `pythinker update`, `pip install --upgrade pythinker-code==0.25.0`, or use the native installer for your platform from the [Releases page](https://github.com/TechMatrix-labs/pythinker-code/releases/latest).
+Upgrade with `pythinker update`, `pip install --upgrade pythinker-code==0.26.0`, or use the native installer for your platform from the [Releases page](https://github.com/TechMatrix-labs/pythinker-code/releases/latest).
 
 
 ---
@@ -214,7 +213,7 @@ matches your OS — no Python, Node, or `uv` prerequisite.
 
 | Platform | Recommended install | Artifact source |
 |---|---|---|
-| **🪟 Windows** | `irm https://pythinker.com/install.ps1 \| iex` | `PythinkerSetup-0.25.0.exe` from [Releases](https://github.com/TechMatrix-labs/pythinker-code/releases/latest) |
+| **🪟 Windows** | `irm https://pythinker.com/install.ps1 \| iex` | `PythinkerSetup-0.26.0.exe` from [Releases](https://github.com/TechMatrix-labs/pythinker-code/releases/latest) |
 | **<img src="https://img.shields.io/badge/-macOS-000000?style=flat-square&logo=apple&logoColor=white" alt="macOS"> / <img src="https://img.shields.io/badge/-Linux-FCC624?style=flat-square&logo=linux&logoColor=black" alt="Linux">** | `curl -fsSL https://pythinker.com/install.sh \| bash` | native tarball from [Releases](https://github.com/TechMatrix-labs/pythinker-code/releases/latest) |
 | **<img src="https://img.shields.io/badge/-macOS-000000?style=flat-square&logo=apple&logoColor=white" alt="macOS"> — Homebrew** | `brew install TechMatrix-labs/pythinker/pythinker-code` | auto-published Homebrew tap |
 | **<img src="https://img.shields.io/badge/-Linux-FCC624?style=flat-square&logo=linux&logoColor=black" alt="Linux"> — system package** | Download the `.deb` or `.rpm` for your distro below | [Releases](https://github.com/TechMatrix-labs/pythinker-code/releases/latest) |
@@ -239,7 +238,7 @@ pythinker                      # start the interactive TUI
 
 ### 🪟 Windows — native installer
 
-`PythinkerSetup-0.25.0.exe` is a signed* Inno Setup wizard. Installs per-user
+`PythinkerSetup-0.26.0.exe` is a signed* Inno Setup wizard. Installs per-user
 into `%LOCALAPPDATA%\Programs\Pythinker`, registers `pythinker` on your user
 PATH (`HKCU\Environment`), broadcasts `WM_SETTINGCHANGE` so new shells see
 the change. **No UAC prompt.**
@@ -250,13 +249,13 @@ irm https://pythinker.com/install.ps1 | iex
 
 # Or manually download the installer + checksum from the Releases page,
 # verify with Get-FileHash, then run:
-.\PythinkerSetup-0.25.0.exe
+.\PythinkerSetup-0.26.0.exe
 
 # Open a fresh PowerShell
 pythinker --version
 ```
 
-**Per-machine install** (IT-managed boxes): `.\PythinkerSetup-0.25.0.exe /ALLUSERS`
+**Per-machine install** (IT-managed boxes): `.\PythinkerSetup-0.26.0.exe /ALLUSERS`
 installs to `%ProgramFiles%\Pythinker` and writes PATH to HKLM (requires admin).
 
 **Upgrade:** `pythinker update` from inside the running app — it downloads
@@ -307,26 +306,26 @@ attached to every GitHub Release.
 
 ```sh
 # Debian / Ubuntu (x86_64)
-sudo dpkg -i pythinker-code_0.25.0_amd64.deb
+sudo dpkg -i pythinker-code_0.26.0_amd64.deb
 sudo apt-get install -f       # only if dpkg reports missing deps
 
 # Debian / Ubuntu (ARM64)
-sudo dpkg -i pythinker-code_0.25.0_arm64.deb
+sudo dpkg -i pythinker-code_0.26.0_arm64.deb
 
 # Fedora / RHEL / openSUSE (x86_64)
-curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.25.0/pythinker-code-0.25.0.x86_64.rpm
-curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.25.0/pythinker-code-0.25.0.x86_64.rpm.sha256
-sha256sum -c pythinker-code-0.25.0.x86_64.rpm.sha256
+curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.26.0/pythinker-code-0.26.0.x86_64.rpm
+curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.26.0/pythinker-code-0.26.0.x86_64.rpm.sha256
+sha256sum -c pythinker-code-0.26.0.x86_64.rpm.sha256
 # Fedora / RHEL:
-sudo dnf install ./pythinker-code-0.25.0.x86_64.rpm
+sudo dnf install ./pythinker-code-0.26.0.x86_64.rpm
 # openSUSE:
-sudo zypper install ./pythinker-code-0.25.0.x86_64.rpm
+sudo zypper install ./pythinker-code-0.26.0.x86_64.rpm
 
 # Fedora / RHEL (aarch64)
-curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.25.0/pythinker-code-0.25.0.aarch64.rpm
-curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.25.0/pythinker-code-0.25.0.aarch64.rpm.sha256
-sha256sum -c pythinker-code-0.25.0.aarch64.rpm.sha256
-sudo dnf install ./pythinker-code-0.25.0.aarch64.rpm
+curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.26.0/pythinker-code-0.26.0.aarch64.rpm
+curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.26.0/pythinker-code-0.26.0.aarch64.rpm.sha256
+sha256sum -c pythinker-code-0.26.0.aarch64.rpm.sha256
+sudo dnf install ./pythinker-code-0.26.0.aarch64.rpm
 ```
 
 Both packages drop a small `/usr/bin/pythinker` launcher that execs the real
@@ -335,8 +334,8 @@ binary under `/usr/lib/pythinker/`, so your `$PATH` stays tidy.
 **Verify before install:**
 
 ```sh
-sha256sum -c pythinker-code_0.25.0_amd64.deb.sha256        # Debian/Ubuntu
-sha256sum -c pythinker-code-0.25.0.x86_64.rpm.sha256       # Fedora/RHEL
+sha256sum -c pythinker-code_0.26.0_amd64.deb.sha256        # Debian/Ubuntu
+sha256sum -c pythinker-code-0.26.0.x86_64.rpm.sha256       # Fedora/RHEL
 ```
 
 **Upgrade:** download the new `.deb`/`.rpm` from Releases and `dpkg -i` /
@@ -366,7 +365,7 @@ at `~/.local/bin/pythinker`.
 curl -fsSL https://pythinker.com/install.sh | bash
 
 # Pin a specific version
-curl -fsSL https://pythinker.com/install.sh | bash -s -- --version 0.25.0
+curl -fsSL https://pythinker.com/install.sh | bash -s -- --version 0.26.0
 
 # Custom prefix (defaults to $HOME/.local)
 curl -fsSL https://pythinker.com/install.sh | bash -s -- --prefix /opt/pythinker

{pythinker_code-0.25.0 → pythinker_code-0.26.0}/README.md

@@ -50,16 +50,15 @@ It speaks the [**Agent Client Protocol (ACP)**](https://github.com/agentclientpr
 
 ---
 
-## 🆕 What's New in 0.25.0
+## 🆕 What's New in 0.26.0
 
-- **`Fetch` re-checks every redirect hop against the SSRF guard.** Redirects were followed without re-validating the destination, so a public URL could redirect to a link-local address (e.g. a cloud metadata endpoint) and slip past the guard. Every hop is now re-validated, closing the public→link-local bypass.
-- **Web domain allowlist for `Fetch` and `Search`.** Set `web.allowed_domains` to restrict which hosts the web tools may reach — `Fetch` (every redirect hop included) and `Search` reject anything off the list. Leave it unset to keep web access unrestricted.
-- **Crash-consistent background tasks.** Task state is serialised under a cross-process lock, terminal updates route through a single finalizer, and recovery reconciles records left divergent by a crash or kill without clobbering a live agent. Bash output is capped (default 50 MiB), kills escalate SIGTERM→SIGKILL, and aged terminal tasks are pruned (default 7 days).
-- **Calmer TUI with live tool feedback.** Tool calls show a "preparing" row during approval, stream shell output as a running tail, and render a live Markdown preview as the model writes. The todo list no longer renders twice mid-turn, links open through a detached launcher, and the terminal is restored to a sane state on `SIGTERM`/`SIGQUIT` and at exit.
-- **Steadier agent loop.** The model is nudged when a turn ends on a bare statement of intent, steered away from blocking on one background task while siblings run, and a `SetTodoList` whose todos arrive as a JSON string is parsed transparently instead of failing validation.
-- **Unified report rendering.** Code review, verify, and security-review output share one standardized renderer, including `report` blocks emitted by skills and agents.
+- **Install and update no longer 404 on a freshly published release.** Releases are now published as a prerelease and promoted to "latest" only once every platform asset has uploaded, and the curl/PowerShell installers wait for their archive + checksum before downloading — so `irm | iex` / `curl | bash` run mid-publish no longer fail on a missing installer.
+- **OpenCode Go model catalog refreshes on startup without a re-login.** Bundled catalog changes (new models, corrected provider shapes) now reach existing installs through the every-startup refresh, across both OpenAI- and Anthropic-shaped providers, while preserving your default model and thinking settings.
+- **Homebrew install shows the logo and fails fast on a missing tap token.** `brew install` now prints the robot-head banner via formula caveats, and the tap workflow reports an actionable error when `HOMEBREW_TAP_TOKEN` is missing instead of an opaque git failure.
+- **Markdown and report rendering hardened against real model output.** Code-span pipes inside tables no longer fracture the table, and a `report` block nested inside a documentation fence is no longer wrongly promoted — report fences are now extracted via a markdown-it AST walk.
+- **Steadier agent editing and a restored update prompt.** File-replace edit handling is hardened, subagent prompt persistence and the agent's working language are preserved, and the blocking pre-start update prompt is restored so it runs before every interactive session.
 
-Upgrade with `pythinker update`, `pip install --upgrade pythinker-code==0.25.0`, or use the native installer for your platform from the [Releases page](https://github.com/TechMatrix-labs/pythinker-code/releases/latest).
+Upgrade with `pythinker update`, `pip install --upgrade pythinker-code==0.26.0`, or use the native installer for your platform from the [Releases page](https://github.com/TechMatrix-labs/pythinker-code/releases/latest).
 
 
 ---
@@ -149,7 +148,7 @@ matches your OS — no Python, Node, or `uv` prerequisite.
 
 | Platform | Recommended install | Artifact source |
 |---|---|---|
-| **🪟 Windows** | `irm https://pythinker.com/install.ps1 \| iex` | `PythinkerSetup-0.25.0.exe` from [Releases](https://github.com/TechMatrix-labs/pythinker-code/releases/latest) |
+| **🪟 Windows** | `irm https://pythinker.com/install.ps1 \| iex` | `PythinkerSetup-0.26.0.exe` from [Releases](https://github.com/TechMatrix-labs/pythinker-code/releases/latest) |
 | **<img src="https://img.shields.io/badge/-macOS-000000?style=flat-square&logo=apple&logoColor=white" alt="macOS"> / <img src="https://img.shields.io/badge/-Linux-FCC624?style=flat-square&logo=linux&logoColor=black" alt="Linux">** | `curl -fsSL https://pythinker.com/install.sh \| bash` | native tarball from [Releases](https://github.com/TechMatrix-labs/pythinker-code/releases/latest) |
 | **<img src="https://img.shields.io/badge/-macOS-000000?style=flat-square&logo=apple&logoColor=white" alt="macOS"> — Homebrew** | `brew install TechMatrix-labs/pythinker/pythinker-code` | auto-published Homebrew tap |
 | **<img src="https://img.shields.io/badge/-Linux-FCC624?style=flat-square&logo=linux&logoColor=black" alt="Linux"> — system package** | Download the `.deb` or `.rpm` for your distro below | [Releases](https://github.com/TechMatrix-labs/pythinker-code/releases/latest) |
@@ -174,7 +173,7 @@ pythinker                      # start the interactive TUI
 
 ### 🪟 Windows — native installer
 
-`PythinkerSetup-0.25.0.exe` is a signed* Inno Setup wizard. Installs per-user
+`PythinkerSetup-0.26.0.exe` is a signed* Inno Setup wizard. Installs per-user
 into `%LOCALAPPDATA%\Programs\Pythinker`, registers `pythinker` on your user
 PATH (`HKCU\Environment`), broadcasts `WM_SETTINGCHANGE` so new shells see
 the change. **No UAC prompt.**
@@ -185,13 +184,13 @@ irm https://pythinker.com/install.ps1 | iex
 
 # Or manually download the installer + checksum from the Releases page,
 # verify with Get-FileHash, then run:
-.\PythinkerSetup-0.25.0.exe
+.\PythinkerSetup-0.26.0.exe
 
 # Open a fresh PowerShell
 pythinker --version
 ```
 
-**Per-machine install** (IT-managed boxes): `.\PythinkerSetup-0.25.0.exe /ALLUSERS`
+**Per-machine install** (IT-managed boxes): `.\PythinkerSetup-0.26.0.exe /ALLUSERS`
 installs to `%ProgramFiles%\Pythinker` and writes PATH to HKLM (requires admin).
 
 **Upgrade:** `pythinker update` from inside the running app — it downloads
@@ -242,26 +241,26 @@ attached to every GitHub Release.
 
 ```sh
 # Debian / Ubuntu (x86_64)
-sudo dpkg -i pythinker-code_0.25.0_amd64.deb
+sudo dpkg -i pythinker-code_0.26.0_amd64.deb
 sudo apt-get install -f       # only if dpkg reports missing deps
 
 # Debian / Ubuntu (ARM64)
-sudo dpkg -i pythinker-code_0.25.0_arm64.deb
+sudo dpkg -i pythinker-code_0.26.0_arm64.deb
 
 # Fedora / RHEL / openSUSE (x86_64)
-curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.25.0/pythinker-code-0.25.0.x86_64.rpm
-curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.25.0/pythinker-code-0.25.0.x86_64.rpm.sha256
-sha256sum -c pythinker-code-0.25.0.x86_64.rpm.sha256
+curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.26.0/pythinker-code-0.26.0.x86_64.rpm
+curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.26.0/pythinker-code-0.26.0.x86_64.rpm.sha256
+sha256sum -c pythinker-code-0.26.0.x86_64.rpm.sha256
 # Fedora / RHEL:
-sudo dnf install ./pythinker-code-0.25.0.x86_64.rpm
+sudo dnf install ./pythinker-code-0.26.0.x86_64.rpm
 # openSUSE:
-sudo zypper install ./pythinker-code-0.25.0.x86_64.rpm
+sudo zypper install ./pythinker-code-0.26.0.x86_64.rpm
 
 # Fedora / RHEL (aarch64)
-curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.25.0/pythinker-code-0.25.0.aarch64.rpm
-curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.25.0/pythinker-code-0.25.0.aarch64.rpm.sha256
-sha256sum -c pythinker-code-0.25.0.aarch64.rpm.sha256
-sudo dnf install ./pythinker-code-0.25.0.aarch64.rpm
+curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.26.0/pythinker-code-0.26.0.aarch64.rpm
+curl -LO https://github.com/TechMatrix-labs/pythinker-code/releases/download/v0.26.0/pythinker-code-0.26.0.aarch64.rpm.sha256
+sha256sum -c pythinker-code-0.26.0.aarch64.rpm.sha256
+sudo dnf install ./pythinker-code-0.26.0.aarch64.rpm
 ```
 
 Both packages drop a small `/usr/bin/pythinker` launcher that execs the real
@@ -270,8 +269,8 @@ binary under `/usr/lib/pythinker/`, so your `$PATH` stays tidy.
 **Verify before install:**
 
 ```sh
-sha256sum -c pythinker-code_0.25.0_amd64.deb.sha256        # Debian/Ubuntu
-sha256sum -c pythinker-code-0.25.0.x86_64.rpm.sha256       # Fedora/RHEL
+sha256sum -c pythinker-code_0.26.0_amd64.deb.sha256        # Debian/Ubuntu
+sha256sum -c pythinker-code-0.26.0.x86_64.rpm.sha256       # Fedora/RHEL
 ```
 
 **Upgrade:** download the new `.deb`/`.rpm` from Releases and `dpkg -i` /
@@ -301,7 +300,7 @@ at `~/.local/bin/pythinker`.
 curl -fsSL https://pythinker.com/install.sh | bash
 
 # Pin a specific version
-curl -fsSL https://pythinker.com/install.sh | bash -s -- --version 0.25.0
+curl -fsSL https://pythinker.com/install.sh | bash -s -- --version 0.26.0
 
 # Custom prefix (defaults to $HOME/.local)
 curl -fsSL https://pythinker.com/install.sh | bash -s -- --prefix /opt/pythinker

{pythinker_code-0.25.0 → pythinker_code-0.26.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "pythinker-code"
-version = "0.25.0"
+version = "0.26.0"
 description = "Pythinker Code is your next CLI agent."
 readme = "README.md"
 requires-python = ">=3.12"

{pythinker_code-0.25.0 → pythinker_code-0.26.0}/src/pythinker_code/CHANGELOG.md

@@ -15,6 +15,19 @@ GitHub Releases page; `0.8.0` is the new starting line.
 
 ## Unreleased
 
+## 0.26.0 (2026-05-30)
+
+### What changed in this release
+
+- **Install and update no longer 404 on a freshly published release.** The 0.24.0 "atomic latest" gate turned out to be dead code — it ran only on a `release: published` event that a token-created release never fires, and did not affect `/releases/latest` anyway. The GitHub Release is now created as a **prerelease**, and a new `promote-release.yml` runs on the tag push, waits for every platform asset to finish uploading, then clears the prerelease flag and marks the release latest — the single point at which a version becomes resolvable by installers and the in-app updater. As defense-in-depth, the curl and PowerShell installers now resolve the newest release that actually carries their platform asset (skipping drafts and prereleases) and wait for the archive plus checksum before downloading, so `irm | iex` / `curl | bash` run during the publish window no longer fail on a missing `PythinkerSetup-<ver>.exe` or archive.
+- **OpenCode Go model catalog refreshes on every startup without a re-login.** Bundled catalog changes — new models, corrected provider shapes — previously reached an existing install only after a manual `pythinker login --opencode-go`. OpenCode Go is now wired into the every-startup `refresh_managed_models` task with its own discovery and a dedicated apply that upserts and prunes across both its OpenAI- and Anthropic-shaped providers while preserving your `default_model` and `default_thinking`. Discovery failures are isolated so they cannot abort other providers' saves; provider `base_url` repairs still require a re-login.
+- **Homebrew install shows the logo and fails fast on a missing tap token.** `brew install` now prints the robot-head banner via a formula `caveats` block — the banner previously lived only in the curl and PowerShell installers, which Homebrew never runs. The tap auto-update workflow now detects an empty `HOMEBREW_TAP_TOKEN` (lost in the org migration, which had frozen the tap at 0.23.0) and fails with an actionable error instead of an opaque git exit-128.
+- **Markdown and report rendering hardened against real model output.** Code-span pipes inside Markdown tables are now protected so a `|` inside backticks no longer fractures the table, and a `report` block nested inside an outer documentation fence is no longer wrongly promoted to a report — report fences are extracted via a markdown-it AST walk instead of a flat regex. Backed by a new TUI markdown/report contract test suite grounded in a real security-scan fixture.
+- **Steadier agent editing and a restored update prompt.** File-replace edit handling is hardened, subagent prompt persistence and the agent's working language are preserved across turns, and the todo list stays aligned during a turn. The blocking pre-start update prompt — wired in 0.24.0 but again left unwired by a later refactor — is restored so it runs before every interactive session.
+- **CI runs required checks on every pull request.** The `check`, `test`, and `release-validate` contexts were path-filtered off docs-only (and `sdks/`/`examples/`-only) PRs, so the required statuses never reported and those PRs stayed BLOCKED under branch protection with no clean override. The `pull_request` path filter is dropped so every PR runs each required context exactly once; the `push` trigger keeps its filter unchanged.
+
+Upgrade with `pythinker update`, `pip install --upgrade pythinker-code==0.26.0`, or use the native installer for your OS (see the README install table).
+
 ## 0.25.0 (2026-05-29)
 
 ### What changed in this release

{pythinker_code-0.25.0 → pythinker_code-0.26.0}/src/pythinker_code/agents/default/system.md

@@ -1,5 +1,9 @@
 You are Pythinker — a think-first software engineering agent running on the user's computer. Before you write code, you read code.
 
+# Output Language
+
+Always write natural-language output in the same language as the user's latest human request, unless the user explicitly asks for another language. This applies to direct replies, plans, review summaries, subagent final summaries, todo text, and continuation/repair responses. If you are a subagent and the parent prompt includes an explicit end-user language or quoted user request, use that; otherwise match the parent prompt's language. Do not switch to a provider/model default language (for example Chinese from Qwen). Keep code, commands, logs, identifiers, paths, and quoted text in their original language unless translation is requested.
+
 Your identity, in order of priority:
 
 1. **Code reviewer.** Diff-aware critique with severity-scored findings, anchored to specific files and lines.
@@ -54,6 +58,8 @@ For any codebase, architecture, debugging, security, performance, planning, or "
 
 `title` is required; `scope`, `note`, `location`, and `body` are optional. `severity` must be one of the five listed values. Order does not matter — the renderer groups by severity (critical first) and derives the summary tally. Put narrative prose outside the block, before or after it.
 
+**Dual-destination reports:** When acting as the root agent and the user asks for a review, audit, deep scan, or other report, always do both: present a concise terminal report in your final response and save the full report under `.pythinker/reports/<descriptive-slug>.md`. Create `.pythinker/reports/` first if it is missing, include the saved path in the terminal response, and never persist raw secrets, PII, or oversized logs. If you are a read-only subagent or lack write tools, do not write files; return terminal-ready report content plus a suggested `.pythinker/reports/...` path so the parent can display and persist it.
+
 # Engineering Discipline
 
 These principles govern every engineering response. They override speed: a slow right answer beats a fast wrong one.
@@ -134,8 +140,6 @@ If the `Shell`, `TaskList`, `TaskOutput`, and `TaskStop` tools are available and
 
 If a foreground tool call or a background agent requests approval, the approval is coordinated through the unified approval runtime and surfaced through the root UI channel. Do not assume approvals are local to a single subagent turn.
 
-When responding to the user, you MUST use the SAME language as the user, unless explicitly instructed to do otherwise.
-
 # General Guidelines for Coding
 
 When building something from scratch, you should:

{pythinker_code-0.25.0 → pythinker_code-0.26.0}/src/pythinker_code/auth/opencode_go.py

@@ -21,6 +21,10 @@ OPENCODE_GO_BASE_URL = "https://opencode.ai/zen/go/v1"
 OPENCODE_GO_ANTHROPIC_BASE_URL = "https://opencode.ai/zen/go"
 OPENCODE_GO_OPENAI_PROVIDER_KEY = "managed:opencode-go-openai"
 OPENCODE_GO_ANTHROPIC_PROVIDER_KEY = "managed:opencode-go-anthropic"
+OPENCODE_GO_PROVIDER_KEYS = (
+    OPENCODE_GO_OPENAI_PROVIDER_KEY,
+    OPENCODE_GO_ANTHROPIC_PROVIDER_KEY,
+)
 OPENCODE_GO_DEFAULT_MODEL_ALIAS = "opencode-go/kimi-k2.6"
 OPENCODE_GO_DEFAULT_CONTEXT = 262_000
 
@@ -297,6 +301,94 @@ async def _discover_opencode_go_models(api_key: str) -> tuple[OpenCodeGoModel, .
     return _build_models(model_ids, metadata)
 
 
+def _opencode_go_api_key(config: Config) -> str | None:
+    """The saved OpenCode Go key (both providers share it), or None if absent."""
+    for provider_key in OPENCODE_GO_PROVIDER_KEYS:
+        provider = config.providers.get(provider_key)
+        if provider is None:
+            continue
+        value = provider.api_key.get_secret_value().strip()
+        if value:
+            return value
+    return None
+
+
+def apply_opencode_go_models(config: Config, models: tuple[OpenCodeGoModel, ...]) -> bool:
+    """Upsert discovered OpenCode Go models and prune ones no longer offered,
+    across both shape providers.
+
+    Unlike ``_apply_opencode_go_config`` (the login path), this preserves the
+    user's ``default_model`` and ``default_thinking`` — it only reassigns the
+    default when the currently selected model was pruned. This mirrors
+    ``platforms._apply_models`` for the two-provider OpenCode Go split, so the
+    background refresh can keep the catalog current without resetting choices.
+
+    Returns True if ``config.models`` (or the default) changed.
+    """
+    changed = False
+    aliases: list[str] = []
+    for model in models:
+        alias = model.alias
+        aliases.append(alias)
+        existing = config.models.get(alias)
+        if existing is None:
+            config.models[alias] = LLMModel(
+                provider=model.provider_key,
+                model=model.model_id,
+                max_context_size=model.max_context_size,
+                display_name=model.display_name,
+            )
+            changed = True
+            continue
+        if existing.provider != model.provider_key:
+            existing.provider = model.provider_key
+            changed = True
+        if existing.model != model.model_id:
+            existing.model = model.model_id
+            changed = True
+        if existing.max_context_size != model.max_context_size:
+            existing.max_context_size = model.max_context_size
+            changed = True
+        if existing.display_name != model.display_name:
+            existing.display_name = model.display_name
+            changed = True
+
+    alias_set = set(aliases)
+    removed_default = False
+    for alias, model_cfg in list(config.models.items()):
+        if model_cfg.provider not in OPENCODE_GO_PROVIDER_KEYS:
+            continue
+        if alias in alias_set:
+            continue
+        del config.models[alias]
+        if config.default_model == alias:
+            removed_default = True
+        changed = True
+
+    if removed_default:
+        config.default_model = aliases[0] if aliases else next(iter(config.models), "")
+        changed = True
+    elif config.default_model and config.default_model not in config.models:
+        config.default_model = next(iter(config.models), "")
+        changed = True
+    return changed
+
+
+async def refresh_opencode_go_models(config: Config) -> tuple[OpenCodeGoModel, ...] | None:
+    """Re-discover the live OpenCode Go model list for the startup managed-model
+    refresh, reusing the saved API key.
+
+    Returns the discovered models to apply, or None when OpenCode Go isn't
+    configured or discovery yields nothing (so callers leave the saved list
+    untouched). Network/HTTP errors propagate to the caller.
+    """
+    api_key = _opencode_go_api_key(config)
+    if api_key is None:
+        return None
+    discovered = await _discover_opencode_go_models(api_key)
+    return discovered or None
+
+
 async def login_opencode_go_api_key(
     config: Config, api_key: str | None = None
 ) -> AsyncIterator[OAuthEvent]:

{pythinker_code-0.25.0 → pythinker_code-0.26.0}/src/pythinker_code/auth/platforms.py

@@ -212,6 +212,13 @@ async def refresh_managed_models(config: Config) -> bool:
     if not config.is_from_default_location:
         return False
 
+    from pythinker_code.auth.opencode_go import (
+        OPENCODE_GO_PROVIDER_KEYS,
+        OpenCodeGoModel,
+        apply_opencode_go_models,
+        refresh_opencode_go_models,
+    )
+
     managed_providers = {
         key: provider for key, provider in config.providers.items() if is_managed_provider_key(key)
     }
@@ -222,6 +229,11 @@ async def refresh_managed_models(config: Config) -> bool:
     updates: list[tuple[str, str, list[ModelInfo]]] = []
     oauth_manager = None
     for provider_key, provider in managed_providers.items():
+        # OpenCode Go uses a two-shape (OpenAI- + Anthropic-compatible) provider
+        # split that the generic single-provider path below can't express, so it
+        # is refreshed via its own discovery after this loop.
+        if provider_key in OPENCODE_GO_PROVIDER_KEYS:
+            continue
         platform_id = parse_managed_provider_key(provider_key)
         if not platform_id:
             continue
@@ -341,12 +353,22 @@ async def refresh_managed_models(config: Config) -> bool:
         if _apply_models(config, provider_key, platform_id, models):
             changed = True
 
+    opencode_go_models: tuple[OpenCodeGoModel, ...] | None = None
+    try:
+        opencode_go_models = await refresh_opencode_go_models(config)
+    except (aiohttp.ClientError, TimeoutError, ValueError) as exc:
+        logger.warning("Failed to refresh OpenCode Go models: {error}", error=exc)
+    if opencode_go_models and apply_opencode_go_models(config, opencode_go_models):
+        changed = True
+
     if changed:
         config_for_save = load_config()
         save_changed = False
         for provider_key, platform_id, models in updates:
             if _apply_models(config_for_save, provider_key, platform_id, models):
                 save_changed = True
+        if opencode_go_models and apply_opencode_go_models(config_for_save, opencode_go_models):
+            save_changed = True
         if save_changed:
             save_config(config_for_save)
     return changed

{pythinker_code-0.25.0 → pythinker_code-0.26.0}/src/pythinker_code/subagents/core.py

@@ -18,6 +18,16 @@ from pythinker_code.subagents.builder import SubagentBuilder
 from pythinker_code.subagents.models import AgentLaunchSpec, AgentTypeDefinition
 from pythinker_code.subagents.store import SubagentStore
 
+SUBAGENT_OUTPUT_LANGUAGE_INSTRUCTION = """\
+<output-language>
+Write natural-language output in the same language as the original user request or task
+prompt, unless that request explicitly asks for another language. Do not switch to a
+model/provider default language (for example Chinese from Qwen). Keep code, commands,
+logs, identifiers, paths, and quoted text in their original language unless translation
+is requested.
+</output-language>
+""".strip()
+
 if TYPE_CHECKING:
     from pythinker_code.soul.agent import Runtime
 
@@ -33,6 +43,12 @@ class SubagentRunSpec:
     resumed: bool
 
 
+def _prepend_output_language_instruction(prompt: str) -> str:
+    if not prompt.strip():
+        return SUBAGENT_OUTPUT_LANGUAGE_INSTRUCTION
+    return f"{SUBAGENT_OUTPUT_LANGUAGE_INSTRUCTION}\n\n{prompt}"
+
+
 async def prepare_soul(
     spec: SubagentRunSpec,
     runtime: Runtime,
@@ -77,6 +93,7 @@ async def prepare_soul(
         git_ctx = await collect_git_context(runtime.builtin_args.PYTHINKER_WORK_DIR)
         if git_ctx:
             prompt = f"{git_ctx}\n\n{prompt}"
+    prompt = _prepend_output_language_instruction(prompt)
 
     # 5. Write prompt snapshot (debugging aid)
     store.prompt_path(spec.agent_id).write_text(prompt, encoding="utf-8")

{pythinker_code-0.25.0 → pythinker_code-0.26.0}/src/pythinker_code/subagents/runner.py

@@ -51,7 +51,9 @@ _SUMMARY_MIN_LENGTH_BY_TYPE: dict[str, int] = {
     "plan": 300,
 }
 SUMMARY_CONTINUATION_PROMPT = """
-Your previous response was too brief. Please provide a more comprehensive summary that includes:
+Your previous response was too brief. Please provide a more comprehensive summary.
+Keep the same natural-language output language as the original task/user request; do not switch
+to a model/provider default language. Include:
 
 1. Specific technical details and implementations
 2. Detailed findings and analysis

{pythinker_code-0.25.0 → pythinker_code-0.26.0}/src/pythinker_code/tools/file/replace.py

@@ -1,8 +1,9 @@
+import json
 from collections.abc import Callable
 from pathlib import Path
-from typing import override
+from typing import Any, cast, override
 
-from pydantic import BaseModel, Field
+from pydantic import BaseModel, Field, model_validator
 from pythinker_core.tooling import CallableTool2, ToolError, ToolReturnValue
 from pythinker_host.path import HostPath
 
@@ -41,6 +42,69 @@ class Params(BaseModel):
         )
     )
 
+    @staticmethod
+    def _json_string_to_edit(value: Any) -> Any:
+        if not isinstance(value, str):
+            return value
+        try:
+            parsed: Any = json.loads(value)
+        except json.JSONDecodeError:
+            return value
+        return (
+            cast(dict[str, Any] | list[Any], parsed) if isinstance(parsed, (dict, list)) else value
+        )
+
+    @classmethod
+    def _normalize_edit_aliases(cls, value: Any) -> Any:
+        value = cls._json_string_to_edit(value)
+        if isinstance(value, list):
+            return [cls._normalize_edit_aliases(item) for item in cast(list[Any], value)]
+        if not isinstance(value, dict):
+            return value
+
+        normalized: dict[str, Any] = dict(cast(dict[str, Any], value))
+        if "old" not in normalized and "oldText" in normalized:
+            normalized["old"] = normalized["oldText"]
+        if "new" not in normalized and "newText" in normalized:
+            normalized["new"] = normalized["newText"]
+        if "replace_all" not in normalized and "replaceAll" in normalized:
+            normalized["replace_all"] = normalized["replaceAll"]
+        return normalized
+
+    @model_validator(mode="before")
+    @classmethod
+    def _normalize_common_edit_shapes(cls, data: Any) -> Any:
+        """Accept common model-generated StrReplaceFile argument shapes.
+
+        Agents occasionally pass the nested ``edit`` payload as a JSON string, or
+        flatten ``old``/``new`` at the top level after seeing the UI label this
+        tool as "Update". Normalize those shapes before Pydantic validates the
+        canonical schema.
+        """
+        if not isinstance(data, dict):
+            return data
+
+        values: dict[str, Any] = dict(cast(dict[str, Any], data))
+        if "edit" in values:
+            values["edit"] = cls._normalize_edit_aliases(values["edit"])
+            return values
+        if "edits" in values:
+            values["edit"] = cls._normalize_edit_aliases(values["edits"])
+            return values
+
+        old_key = "old" if "old" in values else "oldText" if "oldText" in values else None
+        new_key = "new" if "new" in values else "newText" if "newText" in values else None
+        if old_key is not None and new_key is not None:
+            edit: dict[str, Any] = {"old": values[old_key], "new": values[new_key]}
+            if "replace_all" in values:
+                edit["replace_all"] = values["replace_all"]
+            elif "replaceAll" in values:
+                edit["replace_all"] = values["replaceAll"]
+            values["edit"] = edit
+            return values
+
+        return values
+
 
 class StrReplaceFile(CallableTool2[Params]):
     name: str = "StrReplaceFile"

{pythinker_code-0.25.0 → pythinker_code-0.26.0}/src/pythinker_code/ui/shell/__init__.py

@@ -54,6 +54,7 @@ from pythinker_code.ui.shell.slash import SKILL_COMMAND_PREFIX, shell_mode_regis
 from pythinker_code.ui.shell.slash import registry as shell_slash_registry
 from pythinker_code.ui.shell.update import (
     pending_update_notice,
+    prompt_pre_start_update,
     refresh_update_cache_if_due,
 )
 from pythinker_code.ui.shell.visualize import (
@@ -617,9 +618,15 @@ class Shell:
             finally:
                 self._cancel_background_tasks()
 
-        # Start the update check in the background only. Startup must never be
-        # blocked by version polling; users get a cached toast and can run
-        # /update when they are ready.
+        # Blocking pre-start update prompt. Must run before _auto_update so the
+        # same upgrade isn't shown as both a blocking menu and a background
+        # toast; if the user picks "Skip this session" the toast is suppressed
+        # by _skipped_version_this_session. May raise typer.Exit on "Update now"
+        # or "Exit" — that's the documented behavior. prompt_pre_start_update
+        # self-suppresses for source checkouts and non-TTY sessions.
+        await prompt_pre_start_update()
+
+        # Start auto-update background task if not disabled.
         if get_env_bool("PYTHINKER_CLI_NO_AUTO_UPDATE"):
             logger.info("Auto-update disabled by PYTHINKER_CLI_NO_AUTO_UPDATE environment variable")
         else: