A local-first multi-agent app development platform, built on PenguinHarness
The 0.2.5 release post, and the announcement bar entry that links to it (#474)
Yaowei Zhengfa9d3c7765ee
diff --git a/changelog/unreleased/2026-08-27-landing-homepage-refinement.md b/changelog/unreleased/2026-08-27-landing-homepage-refinement.mdindex aa349d64..12526fe2 100644--- a/changelog/unreleased/2026-08-27-landing-homepage-refinement.md+++ b/changelog/unreleased/2026-08-27-landing-homepage-refinement.md@@ -1,7 +1,7 @@# Refine the landing page hierarchy and downloads- **Date:** 2026-08-27-- **Type:** fix+- **Type:** feature- **Scope:** `landing`, `docs`- **PR:** [#473](https://github.com/Prism-Shadow/penguin-harness/pull/473)diff --git a/changelog/unreleased/2026-08-27-landing-homepage-refinement.zh.md b/changelog/unreleased/2026-08-27-landing-homepage-refinement.zh.mdindex eac38440..13472f05 100644--- a/changelog/unreleased/2026-08-27-landing-homepage-refinement.zh.md+++ b/changelog/unreleased/2026-08-27-landing-homepage-refinement.zh.md@@ -1,7 +1,7 @@# 优化官网主页层级与下载体验- **Date:** 2026-08-27-- **Type:** fix+- **Type:** feature- **Scope:** `landing`, `docs`- **PR:** [#473](https://github.com/Prism-Shadow/penguin-harness/pull/473)diff --git a/changelog/unreleased/2026-08-27-landing-page-focus.md b/changelog/unreleased/2026-08-27-landing-page-focus.mddeleted file mode 100644index ea0b226a..00000000--- a/changelog/unreleased/2026-08-27-landing-page-focus.md+++ /dev/null@@ -1,16 +0,0 @@-# Focus the landing and download pages--- **Date:** 2026-08-27-- **Type:** feature-- **Scope:** `landing`-- **PR:** [#473](https://github.com/Prism-Shadow/penguin-harness/pull/473)--[中文版](2026-08-27-landing-page-focus.zh.md)--The landing page was reorganized around the local-first Agent application platform, with clearer desktop and command-line entry points and a shorter product story.--## Details--- Reduced the homepage to product proof, self-evolution, examples, and a desktop-first quick start.-- Added distinct desktop, command-line install, Web UI, and CLI paths.-- Added platform marks and a block-based GitHub and OSS speed report to the download page.diff --git a/changelog/unreleased/2026-08-27-landing-page-focus.zh.md b/changelog/unreleased/2026-08-27-landing-page-focus.zh.mddeleted file mode 100644index 5755c902..00000000--- a/changelog/unreleased/2026-08-27-landing-page-focus.zh.md+++ /dev/null@@ -1,16 +0,0 @@-# 聚焦官网首页与下载页--- **Date:** 2026-08-27-- **Type:** feature-- **Scope:** `landing`-- **PR:** [#473](https://github.com/Prism-Shadow/penguin-harness/pull/473)--[English](2026-08-27-landing-page-focus.md)--官网首页围绕本地优先的 Agent 应用自动开发平台重新组织,以更清晰的桌面端和命令行入口串联更精简的产品叙事。--## 细节--- 首页内容收敛为产品实景、自进化、案例与桌面端优先的快速开始。-- 分别提供桌面端、命令行安装、Web UI 与 CLI 启动路径。-- 下载页增加系统标识,以及以方块呈现的 GitHub 与 OSS 测速结果。diff --git a/packages/landing/README.md b/packages/landing/README.mdindex 640ebbd4..a780428b 100644--- a/packages/landing/README.md+++ b/packages/landing/README.md@@ -37,6 +37,6 @@ excerpt: One-line summary for the list page## Screenshots-`pnpm --filter @prismshadow/penguin-landing shots` regenerates `src/assets/shots/`: the script ships a scripted mock LLM (speaking both Anthropic SSE and OpenAI chat-completions streams), boots the Web service on a temp data root, drives a real "build an Agent app" conversation (tools actually execute in the Workspace), then captures the chat, trace and evaluation pages per UI language (zh/en, separate users) and theme (light/dark) — 12 shots (`<page>-<lang>-<theme>.webp`, re-encoded to WebP in Chromium to keep sizes down). The landing page always shows the shot matching the visitor's language and theme. Prereqs: build skills/core/server/web first, and have Playwright Chromium installed.+`pnpm --filter @prismshadow/penguin-landing shots` regenerates `src/assets/shots/`: the script ships a scripted mock LLM (speaking both Anthropic SSE and OpenAI chat-completions streams), boots the Web service on a temp data root, drives a real "build an Agent app" conversation (tools actually execute in the Workspace), then captures the chat page, its Trace panel (opened as a tab in the bottom dock) and the evaluation page per UI language (zh/en, separate users) and theme (light/dark) — 12 shots (`<page>-<lang>-<theme>.webp`, re-encoded to WebP in Chromium to keep sizes down). The landing page always shows the shot matching the visitor's language and theme. Prereqs: build skills/core/server/web first, and have Playwright Chromium installed.Part of [PenguinHarness](https://github.com/Prism-Shadow/penguin-harness) · Apache-2.0diff --git a/packages/landing/content/blog/penguinharness-0-2-5.en.md b/packages/landing/content/blog/penguinharness-0-2-5.en.mdnew file mode 100644index 00000000..d7cb96bf--- /dev/null+++ b/packages/landing/content/blog/penguinharness-0-2-5.en.md@@ -0,0 +1,109 @@+---+title: "PenguinHarness 0.2.5: two Flash models, a TokenDance gateway, and subagents you can talk to"+date: 2026-08-27+category: news+excerpt: GLM-5.3 Flash and Qwen 3.8 Flash joined the built-in catalog, and GLM-5.3 Flash is listed three times over — once per route that sells it, at three different prices. A TokenDance gateway group arrived with an authorization flow that mints its own key. And an agent's reach grew three ways: a subagent can be steered mid-run, by the model and by you; the `penguin` CLI became a client of the running server, so an agent can drive the install hosting it; and a Session can be bound to a Feishu app or a Telegram bot and answered from the chat.+---++PenguinHarness 0.2.5 is out. The model catalog gained two low-cost Flash models and a whole new gateway, and an agent's reach grew in three directions: the subagents panel turned from a place you watch a child agent into a place you talk to one, the `penguin` CLI became a client of the running server so an agent can drive the harness hosting it, and a Session can be bound to a Feishu app or a Telegram bot and answered from there.++## Two Flash models, one of them three times over++`glm-5.3-flash` and `qwen3.8-flash` joined the built-in catalog. Both carry a million-token context window, both read images, and both price roughly an order of magnitude under their non-Flash siblings.++GLM-5.3 Flash is in the catalog three times, because three routes sell it — Z.AI directly, OpenRouter, and the new TokenDance group — and each row records what that route charges.++++The three prices disagree for two different reasons, and one of them is a convention worth knowing. TokenDance simply charges its own rate. But Z.AI is running a 50% promotion on this model through 2026-09-09, and the two rows treat it differently on purpose: a **direct-vendor row records the vendor's list price**, so the Z.AI row reads $0.03 / $0.15 / $0.50 per million tokens, while a **gateway row records what the gateway actually bills**, so the OpenRouter row reads half that. Until the promotion lapses, the direct row over-states what Z.AI charges you by a factor of two.++The Vision badge on the direct row is newer than it looks. AgentHub 0.4.8 taught the GLM client to forward `image_url` parts — in a prompt and in a tool result alike — and it does so for `glm-5.3-flash` and nothing else. Every other GLM id answers an image with `"GLM <id> does not support image inputs."` rather than quietly dropping it, which is why the rest of that group is still marked vision-off.++Qwen 3.8 Flash sits in the Qwen Pay-As-You-Go group, a million-token input window against a 131K output cap, images included.++++## A TokenDance gateway, and a key you do not have to copy++TokenDance is now a provider group of its own, reaching `https://tokendance.space/gateway/v1` over Chat Completions, with seven presets.++++**Authorize key** is the part worth pointing at. It sends you to TokenDance's own authorization page, mints a key on your account there, and writes it into every model in the group — so a first run no longer starts with a trip to a console and a copy-paste. The PKCE verifier is generated on the server and never sent to a browser, and the minted key goes straight into the group's models without passing through one either; it appears in no response, log line, URL or error message. Where a redirect cannot reach you — the desktop shell, an odd deployment — a manual mode shows a one-time code to paste instead.++The group's prices are **official list prices**, the same convention the two Qwen groups follow, and two of its rows are on live promotions: `glm-5.3-flash` at 50% off and `qwen3.8-max` at 20% off. So for those two, the catalog currently costs you more than the gateway does. Both promotions are tagged in TokenDance's public catalog API, which needs no credential, so whether one still runs is checkable rather than guessable.++Alongside the group, model requests now name PenguinHarness to the gateways that read an app-attribution header — `HTTP-Referer` and `X-OpenRouter-Title` for OpenRouter, `X-App-URL` for TokenDance. The header is chosen by the endpoint's host, not by the provider group, so a model filed under **custom** that points at one of those gateways is attributed the same way.++None of this reaches a Project you already have on its own. Presets are copied into `.project_config.toml` when a Project is created and nothing rewrites them afterwards; **sync presets** on the models page is the one path. It appends what is missing and updates catalog-owned fields on what is there, deletes nothing, and does not touch your stored default model or any credential.++## Subagents you can talk to++A subagent used to be fire-and-observe. `input_subagent` refused a prompt while the child was still running, the panel offered no way to correct or stop one at all, and stopping the main agent left its children running. All of that is gone, replaced by one mechanism: the steering channel you already have into the main session, pointed at the child session, and reachable by the model and by you.++++For the model, `input_subagent` gained a lot. A prompt sent to a running child is injected mid-run as a steering message at the child's next step. A new `abort` argument stops the child's current run only — the session survives for follow-ups — and combining `abort` with a prompt interrupts and redirects in one call. Its output is now the child's most recent complete reply, an idempotent snapshot on every access rather than a delta drain. And a released `subagent_id` revives automatically when the model messages it again.++For you, the selected child is driven by the **same composer as the main conversation**: text, skills and slash commands, a per-turn thinking level, the context ring showing the child's own usage, and the approval-mode selector. A message is a user input on the child whatever state it is in — steering while it runs, a new round while it is idle, and a revival when the session had already ended. "This subagent has ended" is no longer a dead end.++The last gap was approvals. A background child that hit a tool approval used to be auto-denied, because the parent task's end converged every pending approval and a child with no active poll window had nowhere to ask. Now a child approval with nowhere else to go escalates to you as an ordinary approval card, and a parent task ending converges only the main session's approvals — the child's card stays rendered until you decide.++One trade-off comes with it, and it is a real one. Since a subagent session is always resumable and never destroyed, the kill notion for subagents is gone: `kill_subagent` is **removed**, and `kill_command` is folded into `input_command` as `kill: true`. No compatibility is kept for the two removed names — a model calling one gets an unknown-tool failure. See [Upgrading](#upgrading).++## An agent that can drive the harness++`penguin run` and `penguin chat` used to execute a Task in-process against core: no server involved, and nothing the Web App could see. They are HTTP/SSE clients now — the CLI parses arguments and renders the stream, the server runs the Task — so a conversation started from a terminal shows up in the sidebar, and one started in the browser can be picked up from a terminal.++That rebuild is what the rest of this rests on. Beside `run` and `chat` there is now a family for reading and moving an install — `ls`, `input`, `logs`, `agent`, `project`, `cost` and `schedule` — and every command an Agent runs is handed the coordinates to use them. `PENGUIN_API_URL`, `PENGUIN_API_TOKEN`, `PENGUIN_PROJECT_ID`, `PENGUIN_AGENT_ID` and `PENGUIN_SESSION_ID` go into every tool subprocess, so from inside a session a bare `penguin agent ls` reaches the very server running that session, with that Project and that Agent already the defaults. There is no login step anywhere on that path.++++`penguin input` is where this meets the section above: the same verb, one level out. Sent at a **running** session it is absorbed as a course correction at the next step, exactly as a steering message is on a subagent; sent at an idle one it starts a new turn. Without `-m` it polls instead, printing that session's most recent complete assistant reply — an idempotent snapshot, the same semantics `input_subagent` has with an empty prompt. `--timeout` bounds any wait and expires as a soft yield: exit 0, a line naming the follow-up command, and the Task still running on the server.++A session an Agent starts inherits the Agent's own. `run` fills each field left unspecified from the calling session's live values — Workspace, the model pair, approval mode, thinking level — field by field, the same inheritance `run_subagent` applies to children. So from inside an Agent, `penguin run -m "…"` on its own usually does the right thing, and flags are for diverging from it.++Authorization is deliberately plain. The server mints an API token at every boot into `<data-root>/api-token`, mode 0600, and a CLI on the same machine reads it; that token is admin-equivalent, on the grounds that filesystem access to the data root already is admin authority. A remote `--server` gets no such shortcut and is refused without an explicit `PENGUIN_API_TOKEN`.++The recipes ship as a Skill. **`penguin-orchestration`** joined the library's AI App Development group, with no preinstall marker, so a newly created Agent picks it up like the rest of the library. It carries the conventions — orient with the read-only listings before mutating anything, hand bulk material to a new session through a Workspace file rather than through `-m` — and the cautions, which are the real edges: one active Task per session, so a message at a busy session steers it rather than queueing a second; a spawned session inherits your approval mode, so an unattended `always-ask` run just hangs waiting for a human who is not watching; an Agent that can message Agents can build a loop that never terminates; and everything it starts bills the same Project.++## A Session you can reach from Feishu or Telegram++A conversation no longer has to be read in a browser tab. A Session can be bound to a self-built Feishu app or a Telegram bot: messages sent to the bot land in that Session as ordinary user input, and the assistant's replies come back into the chat.++++Both channels sit behind one connector seam, and a Session keeps a saved config for each — Feishu takes an App ID, an App Secret and an API domain, Telegram a single Bot Token — but **at most one of them is enabled**, and the enabled one holds the live connection. Turning the second on is gated with the reason on screen, and refused by the server as well.++Neither channel needs a public URL. Feishu's inbound events arrive over the SDK's WebSocket long connection, Telegram's over an offset-based `getUpdates` long poll — Telegram pushes nothing without a public webhook, so the connector pulls instead. A laptop behind NAT is enough: no tunnel, no webhook endpoint. On connect the poller discards whatever accumulated while nothing was listening rather than replaying a dark period as a flood of Tasks, which is also how Feishu behaves — missed events are simply gone.++Replies arrive as they finish rather than in one block at the end. Each completed assistant message is relayed on its own the moment it completes, so a run that writes working notes between tool calls reaches the chat as that same sequence; the sends of one Session are serialised so they arrive in the order they completed, and chunked under the channels' text limits (Telegram's 4096-character cap is the tightest). In a group chat, the first thing a run sends is threaded onto the message that triggered it and everything after it — including that message's own continuation chunks — is a plain send, so a long answer does not bury the thread in quote blocks.++Credentials follow the models page's interaction: the field always starts empty, a stored secret shows only as its mask, and a "clear stored …" checkbox drops it on save — refused while that channel's connection is enabled, so a live connection can never outrun the credential its store still has. Saving and enabling stay separate concerns, Save writing credentials and the switch connecting or terminating with what is stored. The controls lead the form and the credential fields trail them, which is deliberate: the two channels' field lists differ in length, so controls placed underneath would sit at a different height in each channel and jump on every switch.++One form, two hosts — the session row's menu, and a **Messaging** panel in the conversation's dock beside the Trace panel. A Session holding a live connection carries a small paper plane on its sidebar row, the same mark for every channel, with the channel named in the tooltip and the screen-reader text rather than in the shape. What the model sees is nothing special: inbound text starts a Task as plain composer input, with no marker and no special sender, so it never learns the message came from a chat app. A pending tool approval sends a one-line notice pointing at the web UI, since approving is not something a chat message can do.++## Also in this release++- **The context ring is a button now.** Clicking it breaks the current context into six parts — system prompt, tool definitions, user messages, model messages, tool requests, tool results — as one bar across the whole window, with a dashed mark where compaction will fire and a ranking of the five tools eating the most of it.+- **A red dot leads to the update.** Whenever a software release or an Agent kernel update is available, the chrome carries a phone-style dot, and every dot leads down an unbroken path to the control that performs that update rather than stopping at a notice.+- **Groups go in the order you put them.** Both the chat sidebar's conversation groups and the models page's provider groups can be dragged into a manual order, stored per Project; the chat model picker follows the models page.+- **The sidebar counts by group.** Each group pages its own server stream, so its reveal row names that group's own hidden count and a **Show less** folds it back, and the group list itself paginates ten at a time.+- **The standalone Trace page is gone.** A Trace is read in its conversation's Trace panel, and Trace import moved into System settings with an explicit Project and Agent destination — an imported Trace becomes a listed conversation instead of hiding behind a filter.+- **Agents are easier to start.** The create dialog installs library skills and the skills a chosen project directory already carries, and it can initialize an Agent straight from an exported snapshot package instead of making you create an empty one and import into it.++## Upgrading++Five changes in this release need a decision from you rather than just an update.++**`kill_subagent` and `kill_command` are gone.** Stopping a subagent run is `input_subagent`'s `abort`; terminating a command is `input_command` with `kill: true`. Nothing translates the old names: an Agent whose stored tool configuration names them simply stops assembling those entries, and a model that calls one gets the standard unknown-tool failure. Update anything that pins either name by hand.++**Editing one field of a settings tab freezes that whole tab.** The kernel update's unit is now a settings tab rather than a single config leaf. The upside is that a tab you have customized is never overwritten; the cost is that an Agent with one customized built-in tool stops receiving new built-in tools entirely. "Restore the default configuration" on that tab puts it back under kernel updates.++**An Agent's commands no longer inherit the server's `PENGUIN_*` environment.** No `PENGUIN_*` variable — nor `PORT` or `HOST` — reaches a command an Agent runs, so a harness an Agent starts takes the default data root rather than the serving install's. Outbound proxy variables are the deliberate exception. If a command needs one of those variables, set it in that Agent's vault, which is applied after the host environment is stripped.++**The CLI no longer runs a Task without a server.** `penguin run` and `penguin chat` go through the server API; where none is running they start one locally, and fully offline core-direct execution is no longer a CLI mode — the SDK keeps that capability for embedders. In the same rework the per-user "show CLI sessions" filter was retired and every session is always listed, so Sessions an older CLI left behind now appear in the sidebar on their own. Nothing on disk has to be migrated, and old binaries keep working against their own cores until they are updated.++**New catalog rows are not automatic.** Presets are copied into `.project_config.toml` at Project creation and nothing rewrites them. Use **sync presets** on the models page to pick up GLM-5.3 Flash, Qwen 3.8 Flash and the TokenDance group. A Project older than that group has to sync once before the group — and its **Authorize key** action — appears at all.++Full detail for every change is in [`changelog/0.2.5/`](https://github.com/Prism-Shadow/penguin-harness/tree/main/changelog/0.2.5).diff --git a/packages/landing/content/blog/penguinharness-0-2-5.zh.md b/packages/landing/content/blog/penguinharness-0-2-5.zh.mdnew file mode 100644index 00000000..c5836a20--- /dev/null+++ b/packages/landing/content/blog/penguinharness-0-2-5.zh.md@@ -0,0 +1,109 @@+---+title: "PenguinHarness 0.2.5:两个 Flash 模型、TokenDance 网关,以及可以对话的子智能体"+date: 2026-08-27+category: news+excerpt: GLM-5.3 Flash 与 Qwen 3.8 Flash 进入预置模型库,其中 GLM-5.3 Flash 被列了三次——每条售卖它的通路各一条记录,价格各不相同。新增 TokenDance 网关分组,并带一套能直接签发密钥的授权流程。智能体的触达范围也向三个方向长了一截:子智能体可以在运行途中被纠偏,模型能做,你也能做;`penguin` CLI 变成了服务端的客户端,智能体由此能操作运行着自己的那套 PenguinHarness;Session 还能绑定飞书应用或 Telegram 机器人,直接在聊天里对话。+---++PenguinHarness 0.2.5 发布。预置模型库多了两个低价 Flash 模型和一整个新网关;智能体的触达范围则向三个方向各长了一截:子智能体面板从「看着子智能体跑」变成了「和它对话」,`penguin` CLI 变成了服务端的客户端——这正是智能体得以驱动运行着自己的那套 Harness 的前提,而 Session 可以绑定飞书应用或 Telegram 机器人,从聊天里对话。++## 两个 Flash 模型,其中一个出现了三次++`glm-5.3-flash` 与 `qwen3.8-flash` 进入预置模型库。两者都是百万 Token 上下文窗口,都能读图,价格也都比各自的非 Flash 版本低一个数量级左右。++GLM-5.3 Flash 在模型库里出现了三次,因为有三条通路在卖它——Z.AI 直连、OpenRouter,以及新增的 TokenDance——每条记录写的都是这条通路的收费。++++三个价格不一致有两个原因,其中一个是值得知道的口径差异。TokenDance 只是按自己的费率收费;而 Z.AI 正在对这个模型做 50% 促销,有效期至 2026-09-09——两条记录对这次促销的处理是**有意不同**的:**直连厂商的记录写厂商的官方标价**,所以 Z.AI 那条是每百万 Token $0.03 / $0.15 / $0.50;**网关的记录写网关实际计费的价格**,所以 OpenRouter 那条是它的一半。促销结束之前,直连那条记录把 Z.AI 的实际收费高估了一倍。++直连那条上的「视觉」标记比看上去要新。AgentHub 0.4.8 让 GLM 客户端开始转发 `image_url` 部件——提示词里和工具结果里都转发——而且只对 `glm-5.3-flash` 一个 id 生效。其他每个 GLM id 面对图片都会直接报 `"GLM <id> does not support image inputs."`,而不是悄悄丢掉,这也是该分组其余模型仍标为不支持视觉的原因。++Qwen 3.8 Flash 位于 Qwen Pay-As-You-Go 分组,百万 Token 输入窗口、131K 输出上限,同样支持图片输入。++++## TokenDance 网关,以及一把不用手抄的密钥++TokenDance 现在是独立的供应商分组,通过 Chat Completions 协议访问 `https://tokendance.space/gateway/v1`,随分组提供七个预置模型。++++值得单独点出来的是**自动获取密钥**。它把你送到 TokenDance 自己的授权页,在你的账号下签发一把新密钥,并写入该分组的每一个模型——第一次使用不必再跑一趟控制台手工复制。PKCE 校验串在服务端生成、从不下发给浏览器,签发出的密钥也不经过浏览器直接进入分组的模型配置,不出现在任何响应、日志、URL 或错误信息里。回调到不了你这里的场景(桌面端,或某些部署形态)另有一种手动模式:授权页改为显示一次性验证码,粘贴即可。++该分组的价格记录的是**官方标价**,与下方两个 Qwen 分组同一口径;而其中两条正在促销:`glm-5.3-flash` 打五折、`qwen3.8-max` 打八折。也就是说这两条目前把实际花费算高了。两个促销都在 TokenDance 的公开模型接口里带标记,且该接口无需凭据,所以促销是否还在跑是可以查的,不用猜。++与该分组一同上线的还有应用归属:模型请求现在会向读取该类头部的网关表明自己是 PenguinHarness——OpenRouter 用 `HTTP-Referer` 与 `X-OpenRouter-Title`,TokenDance 用 `X-App-URL`。头部按端点的 host 选择,而非按供应商分组,因此一个归在 **custom** 下、却指向这两个网关的模型,会得到同样的归属标记。++以上这些都不会自动出现在你已有的 Project 里。预置模型是在创建 Project 时复制进 `.project_config.toml` 的,此后不再被改写;模型页的**同步预置**是唯一入口。它只补入缺失项、更新 catalog 所属字段,不删除任何东西,也不会动你保存的默认模型和任何凭据。++## 可以对话的子智能体++子智能体过去是「发出去,然后只能看着」。子会话还在运行时 `input_subagent` 会拒绝新的提示词;面板里既不能纠偏也不能停止;停掉主智能体,它的子会话还在跑。这些现在都没有了,取而代之的是同一套机制:把你本来就有的、对主会话的纠偏通道指向子会话,模型和你都能用。++++对模型来说,`input_subagent` 增强了不少。发给运行中子会话的提示词会作为纠偏消息在它下一步注入。新增的 `abort` 参数只停掉子会话**当前这一轮**——会话本身保留,还能继续对话——`abort` 与提示词一起给出,就是一次调用完成「打断并改道」。它的返回值现在是子会话**最近一条完整回复**,每次访问都是幂等快照,而不是增量抽取。已释放的 `subagent_id` 在模型再次给它发消息时会自动复活。++对你来说,选中的子会话由**与主对话完全相同的输入框**驱动:正文、技能与斜杠指令、按轮设置的思考档位、显示子会话自身用量的上下文环,以及审批模式选择器。无论子会话处于什么状态,你发的都是一条对它的用户输入——运行中是纠偏,空闲时是新一轮,已经结束的则被复活。「该子智能体已结束」不再是死路。++最后一处缺口是审批。后台子会话碰到工具审批过去会被自动拒绝:父任务结束时会把所有待审批一并收敛,而没有活动轮询窗口的子会话根本没地方发问。现在,无处可去的子会话审批会以普通审批卡片的形式升级给你;父任务结束时也只收敛主会话的审批——子会话的卡片会一直留在那里,等你决定。++随之而来的是一处取舍,而且是实打实的。既然子会话始终可恢复、从不销毁,「kill」这个概念对子智能体就不再成立:`kill_subagent` 已**移除**,`kill_command` 并入 `input_command`,改为 `kill: true`。这两个旧名字没有保留任何兼容——模型调用它们会得到标准的「未知工具」失败。详见下方[升级须知](#升级须知)。++## 能驱动 Harness 自己的智能体++`penguin run` 与 `penguin chat` 过去是在进程内直接调用 core 执行任务的:不经过服务端,Web App 也看不见。现在它们是 HTTP/SSE 客户端——CLI 负责解析参数与渲染流式输出,任务本身在服务端执行——因此终端里发起的对话会出现在侧边栏里,浏览器里发起的对话也能从终端接着聊。++这次重建是后面这些的前提。除 `run` 与 `chat` 外,现在还有一组用于查看和操作一套部署的命令——`ls`、`input`、`logs`、`agent`、`project`、`cost` 与 `schedule`——而 Agent 运行的每一条命令都会被交给使用它们所需的坐标。`PENGUIN_API_URL`、`PENGUIN_API_TOKEN`、`PENGUIN_PROJECT_ID`、`PENGUIN_AGENT_ID` 与 `PENGUIN_SESSION_ID` 会注入每一个工具子进程,因此在会话内部,一句不带任何参数的 `penguin agent ls` 访问的正是承载该会话的那个服务端,并且已经以该 Project 与该 Agent 为默认值。这条路径上没有任何登录步骤。++++`penguin input` 正是它与上一节交汇的地方:同一个动作,只是往外挪了一层。发给**运行中**会话时,它会在下一步作为纠偏被吸收,与对子智能体发送纠偏消息完全一致;发给空闲会话则开启新一轮。不带 `-m` 时它转为轮询,打印该会话最近一条完整的助手回复——一份幂等快照,与 `input_subagent` 传空提示词时的语义相同。`--timeout` 为任何等待设定上限,到期是一次软让出:退出码 0、一行点名后续命令的提示,而任务仍在服务端继续跑。++智能体启动的会话会继承它自己的配置。`run` 会把你没有指定的每个字段,逐个从调用方会话的实时取值里补齐——工作区、模型对、审批模式、思考档位——与 `run_subagent` 对子会话的继承是同一套。因此在 Agent 内部,单独一句 `penguin run -m "…"` 通常就是对的,参数是用来偏离它的。++授权模型有意保持朴素。服务端每次启动都会签发一把 API 令牌,写入 `<data-root>/api-token`(权限 0600),同机的 CLI 直接读取它;这把令牌等同管理员权限,理由是能读写数据根的文件系统访问权本来就等同管理员权限。指向远程的 `--server` 没有这层便利,未显式提供 `PENGUIN_API_TOKEN` 时会被拒绝。++这些用法以 Skill 的形式随产品发布。**`penguin-orchestration`** 进入技能库的「AI 应用开发」分组,且没有预装标记,因此新建的 Agent 会像装其余技能一样装上它。它写明了惯例——动手改动之前先用只读列表命令摸清现状,大段材料通过工作区文件交给新会话,而不是塞进 `-m`——也写明了注意事项,那才是真正的边界:每个会话同时只有一个活动任务,因此发给繁忙会话的消息是纠偏而非排入第二个任务;派生的会话继承你的审批模式,因此无人值守的 `always-ask` 会话只会一直挂着,等一个并没有在看的人;能给智能体发消息的智能体,也能搭出一个永不终止的循环;以及它启动的一切都算在同一个 Project 的账上。++## 可以从飞书或 Telegram 联系的 Session++对话不再只能在浏览器标签页里读。Session 可以绑定到一个自建的飞书应用或一个 Telegram 机器人:发给机器人的消息会作为普通用户输入进入该 Session,助手的回复也会回到聊天里。++++两个渠道位于同一层连接器抽象之后,Session 为每个渠道各保存一份配置——飞书需要 App ID、App Secret 与 API 域名,Telegram 只需要一个 Bot Token——但**同时至多只有一个处于启用状态**,持有实时连接的就是它。想启用第二个时,开关会带着原因被拦住,服务端也会拒绝。++两个渠道都不需要公网 URL。飞书的入站事件走 SDK 的 WebSocket 长连接;Telegram 没有公网 webhook 就不会推送,因此连接器改为主动拉取,走基于 offset 的 `getUpdates` 长轮询。一台位于 NAT 之后的笔记本就够了——不必打隧道,也不必准备 webhook 端点。连接时,轮询会把无人连接期间积压的更新直接跳过,而不是把这段空窗期重放成一堆任务;飞书同理——错过的事件就是错过了。++回复是完成一条发一条,而不是在整轮结束时发一大块。每条完整的助手消息在完成的那一刻单独转发出去,因此一轮里在工具调用之间写下的工作笔记,到了聊天里也是同样的一串消息;同一个 Session 的发送被串行化,到达顺序与完成顺序一致,并按各渠道的文本长度上限分块(其中 Telegram 的 4096 字符上限最紧)。在群聊里,一轮中发出的第一段内容会挂在触发它的那条消息下形成话题回复,其后的一切——包括这条消息自身的后续分块——都是普通发送,这样长回答不会把话题埋进一堆引用块里。++凭据沿用模型页的那套交互:输入框永远从空开始,已保存的密钥只以掩码形式显示,旁边的「清除已保存的……」勾选框在保存时将其丢弃——而该渠道的连接处于启用状态时,清除会被拒绝,这样实时连接永远不会跑在它的凭据存量之前。保存与启用是两件独立的事:保存只写凭据,开关则用已存的凭据去连接或断开。表单把控件放在前面、凭据字段排在后面,这是有意为之:两个渠道的字段数量不同,控件若放在字段下方,在两个渠道里的高度就不一致,每次切换都会跳动。++一套表单,两个宿主——会话行的菜单,以及对话面板栏里紧邻 Trace 面板的**消息软件**面板。持有实时连接的 Session 会在侧栏行上带一个小小的纸飞机,所有渠道共用同一个标记,渠道本身由提示气泡与无障碍文本点名,而不是由图形区分。而模型看到的一切并无特别:入站文本以普通输入框内容的形式启动任务,没有任何标记、也没有特殊发送者,因此它不会知道这条消息来自聊天软件。遇到待审批的工具调用时,会发出一行提示指向 Web 界面——审批不是一条聊天消息能完成的事。++## 这一版还有++- **上下文环现在是个按钮。** 点开它,当前上下文被拆成六部分——系统提示词、工具定义、用户消息、模型消息、工具请求、工具结果——画成横跨整个窗口的一条,虚线标出压缩将在何处触发,并列出占用最多的五个工具。+- **一个红点带你走到更新本身。** 只要有可用的软件更新或 Agent 内核更新,界面上就会出现手机式的小红点;每个红点都沿着一条不断的路径通向真正执行更新的那个控件,而不是停在一句通知上。+- **分组按你排的顺序来。** 对话侧栏的会话分组与模型页的供应商分组都可以拖成手动顺序,按 Project 保存;对话的模型选择器跟随模型页的顺序。+- **侧栏按分组各算各的。** 每个分组分页读取自己的服务端流,展开行写的是该分组自己的隐藏数量,并配有**收起**;分组列表本身也按每页十个分页。+- **独立的轨迹页面已移除。** 轨迹在其所属对话的轨迹面板里查看;轨迹导入移入系统设置,并需明确指定目标 Project 与 Agent——导入的轨迹会成为一条正常列出的对话,而不是藏在过滤器后面。+- **新建 Agent 更省事。** 创建弹窗可以一次装好技能库里的多个技能,以及所选项目目录本来就带的技能;也可以直接用导出的快照包初始化一个 Agent,不必先建空 Agent 再导入。++## 升级须知++这一版有五处需要你做一次判断,而不是升级完就没事了。++**`kill_subagent` 与 `kill_command` 已移除。** 停止子会话当前这一轮用 `input_subagent` 的 `abort`;终止命令用 `input_command` 加 `kill: true`。旧名字没有任何转换:Agent 存量工具配置里若点名了它们,这些条目会直接装配不出来;模型调用它们则得到标准的「未知工具」失败。请把手工点名旧名字的地方一并改掉。++**改动设置标签页里的任何一个字段,会冻结整个标签页。** 内核更新的粒度从单个配置叶子改成了一个设置标签页。好处是你自定义过的标签页永远不会被覆盖;代价是只要一个内置工具被改过,该 Agent 就完全收不到新增的内置工具了。在该标签页上「恢复默认配置」,即可让它重新接受内核更新。++**Agent 运行的命令不再继承服务端的 `PENGUIN_*` 环境变量。** 任何 `PENGUIN_*` 变量(以及 `PORT`、`HOST`)都不会传到 Agent 运行的命令里,因此由 Agent 启动的 Harness 会使用默认数据根,而不是当前服务实例的数据根。出网代理相关的变量是有意保留的例外。命令确实需要某个变量时,请写进该 Agent 的 vault——它在剥离宿主环境之后再叠加。++**CLI 不再脱离服务端执行任务。** `penguin run` 与 `penguin chat` 都经由服务端 API;本机没有服务端在跑时,它们会自行拉起一个,而「完全离线、直连 core」不再是 CLI 的一种运行方式——这项能力仍由 SDK 为嵌入方保留。同一次重建中,按用户保存的「显示 CLI 会话」过滤开关被取消,所有会话一律列出,因此旧版 CLI 留下的 Session 现在会自行出现在侧边栏里。磁盘上没有任何数据需要迁移,旧版二进制在更新之前也仍能用它自带的 core 正常工作。++**模型库的新条目不会自动出现。** 预置模型是在创建 Project 时复制进 `.project_config.toml` 的,此后不再被改写。在模型页点**同步预置**即可补入 GLM-5.3 Flash、Qwen 3.8 Flash 与 TokenDance 分组。比该分组更早创建的 Project,需要先同步一次,这个分组——以及它的**自动获取密钥**——才会出现。++每一项改动的完整细节见 [`changelog/0.2.5/`](https://github.com/Prism-Shadow/penguin-harness/tree/main/changelog/0.2.5)。diff --git a/packages/landing/scripts/capture-shots.mjs b/packages/landing/scripts/capture-shots.mjsindex 7af362b4..0ba5d1ac 100644--- a/packages/landing/scripts/capture-shots.mjs+++ b/packages/landing/scripts/capture-shots.mjs@@ -5,8 +5,9 @@* SSE, so whichever client AgentHub routes to gets a valid stream) -> start the Web* server against a temp data root serving the built web dist -> drive a genuine* "build an Agent app" conversation (tools actually execute in the workspace) ->- * screenshot chat / trace view / evaluation center, per UI language (zh / en, each- * with its own user so sidebars stay monolingual) and per theme (light / dark), into+ * screenshot the chat page, its Trace panel and the evaluation center, per UI language+ * (zh / en, each with its own user so sidebars stay monolingual) and per theme+ * (light / dark), into* src/assets/shots/ as <page>-<lang>-<theme>.webp (12 files, re-encoded to WebP* inside Chromium to keep the repo small).*@@ -26,7 +27,9 @@ const ROOT = path.resolve(HERE, "../../..");const OUT_DIR = path.resolve(HERE, "../src/assets/shots");const MOCK_PORT = 8941;const SRV_PORT = 8940;-const BASE = `http://127.0.0.1:${SRV_PORT}`;+// `localhost`, not `127.0.0.1`: that literal address is reserved as the Workspace-preview+// host, and the API answers 401 on it.+const BASE = `http://localhost:${SRV_PORT}`;// Pins the otherwise-random seeded admin password (PENGUIN_SEED_ADMIN_PASSWORD below).const ADMIN_PASSWORD = "penguin-0000";const MOCK = `http://127.0.0.1:${MOCK_PORT}`;@@ -488,6 +491,9 @@ async function provisionUser(adminCookie, lang) {{provider: "deepseek",modelId: "deepseek-v4-pro",+ // The mock speaks Anthropic Messages and OpenAI Chat Completions; pinning the+ // protocol keeps it off the Responses path AgentHub would otherwise pick.+ clientType: "openai-chat",apiKey: "sk-demo",baseUrl: MOCK,contextWindow: 1000000,@@ -605,6 +611,12 @@ try {([t, l]) => {localStorage.setItem("penguin.theme", t);localStorage.setItem("penguin.lang", l);+ // Both docks start closed (no scopes), so the chat shot is unaffected; the+ // ratio only decides how tall the bottom dock is once the Trace tab opens it.+ localStorage.setItem(+ "penguin.dock.layout",+ JSON.stringify({ scopes: {}, bottomRatio: 0.62 }),+ );},[theme, lang],);@@ -628,14 +640,21 @@ try {await page.waitForTimeout(2000);await saveWebp(await page.screenshot(), `chat-${lang}-${theme}.webp`);- // Trace view: the product's canonical deep link carries BOTH agentId and- // sessionId (?sessionId= alone is ignored by TracesPage's focus wiring), and- // auto-selects the Session once its trace list loads. Waiting for the- // execution timeline's exec_command lanes guarantees every language captures- // the same opened trace — stats + a timeline with tool calls — never the- // empty "select a Session" state.- await page.goto(`${BASE}/traces?agentId=default_agent&sessionId=${sessionId}`);- await page.getByText("exec_command").first().waitFor({ timeout: 20000 });+ // Trace view: a Trace is read in its own conversation's Trace panel, opened as a+ // tab in the bottom dock (the toolbar pulls the empty dock open, the picker puts+ // the tab in it). Waiting for a timeline segment titled after the conversation's+ // exec_command call guarantees every language captures the same opened Trace —+ // stats plus a timeline with tool calls — never a still-loading skeleton.+ await page.getByTestId("dock-toggle-bottom").click();+ const bottomDock = page.locator("[data-testid='dock'][data-position='bottom']");+ await bottomDock.getByTestId("dock-pick-trace").click();+ await bottomDock+ .locator('[data-tab-id="trace"][data-active="true"]')+ .waitFor({ timeout: 20000 });+ await bottomDock+ .getByTitle(/exec_command/)+ .first()+ .waitFor({ timeout: 30000 });await page.waitForTimeout(2000);await saveWebp(await page.screenshot(), `traces-${lang}-${theme}.webp`);diff --git a/packages/landing/src/assets/shots/benchmark-en-dark.webp b/packages/landing/src/assets/shots/benchmark-en-dark.webpindex e6229e80..921988f6 100644Binary files a/packages/landing/src/assets/shots/benchmark-en-dark.webp and b/packages/landing/src/assets/shots/benchmark-en-dark.webp differdiff --git a/packages/landing/src/assets/shots/benchmark-en-light.webp b/packages/landing/src/assets/shots/benchmark-en-light.webpindex d8c77e45..80a59ea4 100644Binary files a/packages/landing/src/assets/shots/benchmark-en-light.webp and b/packages/landing/src/assets/shots/benchmark-en-light.webp differdiff --git a/packages/landing/src/assets/shots/benchmark-zh-dark.webp b/packages/landing/src/assets/shots/benchmark-zh-dark.webpindex 23e7dc44..b6c25f18 100644Binary files a/packages/landing/src/assets/shots/benchmark-zh-dark.webp and b/packages/landing/src/assets/shots/benchmark-zh-dark.webp differdiff --git a/packages/landing/src/assets/shots/benchmark-zh-light.webp b/packages/landing/src/assets/shots/benchmark-zh-light.webpindex 37313354..d399c194 100644Binary files a/packages/landing/src/assets/shots/benchmark-zh-light.webp and b/packages/landing/src/assets/shots/benchmark-zh-light.webp differdiff --git a/packages/landing/src/assets/shots/chat-en-dark.webp b/packages/landing/src/assets/shots/chat-en-dark.webpindex 10a40dc8..fcc08110 100644Binary files a/packages/landing/src/assets/shots/chat-en-dark.webp and b/packages/landing/src/assets/shots/chat-en-dark.webp differdiff --git a/packages/landing/src/assets/shots/chat-en-light.webp b/packages/landing/src/assets/shots/chat-en-light.webpindex 767d90da..5f51ed73 100644Binary files a/packages/landing/src/assets/shots/chat-en-light.webp and b/packages/landing/src/assets/shots/chat-en-light.webp differdiff --git a/packages/landing/src/assets/shots/chat-zh-dark.webp b/packages/landing/src/assets/shots/chat-zh-dark.webpindex 1eabeecf..0748ec6a 100644Binary files a/packages/landing/src/assets/shots/chat-zh-dark.webp and b/packages/landing/src/assets/shots/chat-zh-dark.webp differdiff --git a/packages/landing/src/assets/shots/chat-zh-light.webp b/packages/landing/src/assets/shots/chat-zh-light.webpindex ae3a858a..e169e28a 100644Binary files a/packages/landing/src/assets/shots/chat-zh-light.webp and b/packages/landing/src/assets/shots/chat-zh-light.webp differdiff --git a/packages/landing/src/assets/shots/traces-en-dark.webp b/packages/landing/src/assets/shots/traces-en-dark.webpindex 8f735126..6823e9ec 100644Binary files a/packages/landing/src/assets/shots/traces-en-dark.webp and b/packages/landing/src/assets/shots/traces-en-dark.webp differdiff --git a/packages/landing/src/assets/shots/traces-en-light.webp b/packages/landing/src/assets/shots/traces-en-light.webpindex dd81aedd..3cd40704 100644Binary files a/packages/landing/src/assets/shots/traces-en-light.webp and b/packages/landing/src/assets/shots/traces-en-light.webp differdiff --git a/packages/landing/src/assets/shots/traces-zh-dark.webp b/packages/landing/src/assets/shots/traces-zh-dark.webpindex 20855c26..2d4ec7dd 100644Binary files a/packages/landing/src/assets/shots/traces-zh-dark.webp and b/packages/landing/src/assets/shots/traces-zh-dark.webp differdiff --git a/packages/landing/src/assets/shots/traces-zh-light.webp b/packages/landing/src/assets/shots/traces-zh-light.webpindex a4c69031..5f624823 100644Binary files a/packages/landing/src/assets/shots/traces-zh-light.webp and b/packages/landing/src/assets/shots/traces-zh-light.webp differdiff --git a/packages/landing/src/components/announcement-bar.tsx b/packages/landing/src/components/announcement-bar.tsxindex bf2327b4..d453eba2 100644--- a/packages/landing/src/components/announcement-bar.tsx+++ b/packages/landing/src/components/announcement-bar.tsx@@ -20,7 +20,7 @@ const ROTATE_MS = 6000;* freshest news goes at the front.*/const ITEMS = [- { key: "k3AndFree", to: "/blog/free-models-in-penguin-harness" },+ { key: "flashModels", to: "/blog/penguinharness-0-2-5" },{ key: "fireworks", to: "/blog/fireworks-credits-amd" },] as const;@@ -32,7 +32,7 @@ function prefersReducedMotion(): boolean {export function AnnouncementBar() {const texts: Record<(typeof ITEMS)[number]["key"], string> = {- k3AndFree: S.announcement.k3AndFree,+ flashModels: S.announcement.flashModels,fireworks: S.announcement.fireworks,};// pos runs 0..ITEMS.length where ITEMS.length is the clone of slide 0.diff --git a/packages/landing/src/lib/strings-en.ts b/packages/landing/src/lib/strings-en.tsindex df9f38d6..c5a982bc 100644--- a/packages/landing/src/lib/strings-en.ts+++ b/packages/landing/src/lib/strings-en.ts@@ -12,7 +12,7 @@ export const en: Strings = {label: "Announcements",prev: "Previous announcement",next: "Next announcement",- k3AndFree: "Kimi K3 and free models like Ling 3.0 Flash are now available in PenguinHarness",+ flashModels: "GLM-5.3 Flash and Qwen 3.8 Flash are now available in PenguinHarness",fireworks: "Claim $50 in Fireworks API credits with the AMD Developer Program",},diff --git a/packages/landing/src/lib/strings.ts b/packages/landing/src/lib/strings.tsindex 5bfa7d88..8e4b7a21 100644--- a/packages/landing/src/lib/strings.ts+++ b/packages/landing/src/lib/strings.ts@@ -13,7 +13,7 @@ export const zh = {label: "公告",prev: "上一条公告",next: "下一条公告",- k3AndFree: "Kimi K3 与 Ling 3.0 Flash 等免费模型现已在 PenguinHarness 可用",+ flashModels: "GLM-5.3 Flash 与 Qwen 3.8 Flash 现已在 PenguinHarness 可用",fireworks: "携手 AMD 开发者计划:$50 Fireworks API 额度免费领取中",},diff --git a/packages/landing/test/blog.test.ts b/packages/landing/test/blog.test.tsindex 400a5b95..f77f4d50 100644--- a/packages/landing/test/blog.test.ts+++ b/packages/landing/test/blog.test.ts@@ -96,7 +96,7 @@ describe("frontmatter mapping (author / pinned / category)", () => {it("reads the pinned flag and sorts the pinned post first", () => {for (const locale of ["en", "zh"] as const) {const posts = postsFor(locale);- expect(posts.length).toBe(18);+ expect(posts.length).toBe(19);// The launch post stays the single pinned post; newer posts sort under it by date.expect(posts.filter((p) => p.pinned).map((p) => p.slug)).toEqual(["introducing-penguinharness",@@ -120,6 +120,7 @@ describe("frontmatter mapping (author / pinned / category)", () => {it("filters by the news category, newest first", () => {expect(postsFor("en", "news").map((p) => p.slug)).toEqual(["introducing-penguinharness",+ "penguinharness-0-2-5","penguinharness-0-2-4","penguinharness-0-2-3","penguinharness-0-2-2",