Skip to content

feat(super-agent): drop model picker and undocument model flags #7

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Sodawyx merged 1 commit into from
Apr 25, 2026
Merged

feat(super-agent): drop model picker and undocument model flags #7

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 0 additions & 6 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -88,9 +88,6 @@ Credentials land in `~/.agentrun/config.json` under the `default` profile. Use

```bash
$ ar super-agent run --prompt "You are a Python expert"
Loading model services...
? Select model service: svc-tongyi
? Select model: qwen-max
Creating super agent: super-agent-tmp-20260420213045 ...
Ready. Type your message (/help for commands).

Expand Down Expand Up @@ -128,9 +125,6 @@ metadata:
description: "My personal assistant"
spec:
prompt: "You are my helpful assistant"
model:
service: svc-tongyi
name: qwen-max
tools:
- mcp-time-sa
skills: []
Expand Down
6 changes: 0 additions & 6 deletions README_zh.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -87,9 +87,6 @@ ar config set region cn-hangzhou

```bash
$ ar super-agent run --prompt "你是一个 Python 专家"
Loading model services...
? Select model service: svc-tongyi
? Select model: qwen-max
Creating super agent: super-agent-tmp-20260420213045 ...
Ready. Type your message (/help for commands).

Expand Down Expand Up @@ -127,9 +124,6 @@ metadata:
description: "我的助手"
spec:
prompt: "你是我的得力助手"
model:
service: svc-tongyi
name: qwen-max
tools:
- mcp-time-sa
skills: []
Expand Down
28 changes: 5 additions & 23 deletions docs/en/super-agent.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -59,8 +59,6 @@ ar sa run [options]
|------|------|----------|---------|-------------|
| `--name` | string | no | `super-agent-tmp-<YYYYMMDDHHMMSS>` | Agent name. Explicit name → persistent; auto-name → still persistent but clearly temporary. |
| `--prompt`, `-p` | string | no | `You are a helpful assistant.` | System prompt. |
| `--model-service` | string | no | TTY-picker | ModelService name. Required if stdin is not a TTY or `--no-input` is set. |
| `--model` | string | no | TTY-picker | Model name within the ModelService. |
| `--tool` | multi | no | | Tool name, repeatable. |
| `--skill` | multi | no | | Skill name, repeatable. |
| `--sandbox` | multi | no | | Sandbox name, repeatable. |
Expand All @@ -69,23 +67,21 @@ ar sa run [options]
| `--message`, `-m` | string | no | | Initial message — sent right after entering the REPL. |
| `--raw` | flag | no | false | Force raw SSE JSON-line output. |
| `--text-only` | flag | no | false | Only show assistant text (hide tool calls). |
| `--no-input` | flag | no | false | Disable interactive pickers; any missing required arg fails the command. |
| `--no-input` | flag | no | false | Deprecated, no-op. Kept for backward script compatibility. |

### Examples

```bash
# Zero-config — CLI picks ModelService/Model interactively
# Zero-config — server picks a default model
ar sa run

# Explicit prompt and an initial message
ar sa run -p "You write concise Python" -m "Implement FizzBuzz"

# Non-interactive (scripts / CI)
# With tools enabled
ar sa run \
--model-service svc-tongyi --model qwen-max \
--prompt "You are an assistant" \
--tool mcp-time-sa \
--no-input
--tool mcp-time-sa

# Name it and keep it around
ar sa run --name my-helper -p "You are my helper"
Expand Down Expand Up @@ -232,8 +228,6 @@ ar sa create --name <name> [options]
| `--name` | string | yes | Agent name (globally unique). |
| `--description` | string | no | Description. |
| `--prompt`, `-p` | string | no | System prompt. |
| `--model-service` | string | no | ModelService name. |
| `--model` | string | no | Model name. |
| `--tool` | multi | no | Tool name, repeatable. |
| `--skill` | multi | no | Skill name, repeatable. |
| `--sandbox` | multi | no | Sandbox name, repeatable. |
Expand All @@ -244,12 +238,10 @@ ar sa create --name <name> [options]

```bash
ar sa create --name my-helper \
-p "You are my assistant" \
--model-service svc-tongyi --model qwen-max
-p "You are my assistant"

ar sa create --name researcher \
-p "Deep research assistant" \
--model-service svc-tongyi --model qwen-max \
--tool web-search --tool mcp-time-sa \
--skill data-analyzer --skill report-generator
```
Expand Down Expand Up @@ -311,8 +303,6 @@ ar sa update <NAME> [options]
| `NAME` | positional | Agent to update. |
| `--description` | string | New description. |
| `--prompt`, `-p` | string | New prompt. |
| `--model-service` | string | New ModelService. |
| `--model` | string | New model name. |
| `--tool` | multi | Replacement tool list. |
| `--skill` | multi | Replacement skill list. |
| `--sandbox` | multi | Replacement sandbox list. |
Expand All @@ -331,7 +321,6 @@ both fails with exit code `2`.

```bash
ar sa update my-helper -p "You are a concise helper"
ar sa update my-helper --model-service svc-openai --model gpt-4o
ar sa update my-helper --tool web-search --tool calculator
ar sa update my-helper --clear-tools
```
Expand Down Expand Up @@ -405,9 +394,6 @@ metadata:
spec:
prompt: | # optional
You are a helpful assistant.
model: # optional (but required to actually invoke)
service: svc-tongyi
name: qwen-max
tools:
- mcp-time-sa
skills: []
Expand All @@ -423,8 +409,6 @@ spec:
| `metadata.name` | `name` |
| `metadata.description` | `description` |
| `spec.prompt` | `prompt` |
| `spec.model.service` | `model_service_name` |
| `spec.model.name` | `model_name` |
| `spec.tools` | `tools` |
| `spec.skills` | `skills` |
| `spec.sandboxes` | `sandboxes` |
Expand All @@ -442,15 +426,13 @@ metadata:
name: doc-writer
spec:
prompt: "You write clear docs"
model: { service: svc-tongyi, name: qwen-max }
---
apiVersion: agentrun/v1
kind: SuperAgent
metadata:
name: code-reviewer
spec:
prompt: "You are a Python reviewer"
model: { service: svc-tongyi, name: qwen-max }
```

`apply -f` processes documents in order; if any fails, already-succeeded agents
Expand Down
28 changes: 5 additions & 23 deletions docs/zh/super-agent.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -58,8 +58,6 @@ ar sa run [options]
|------|------|------|------|------|
| `--name` | string | 否 | `super-agent-tmp-<YYYYMMDDHHMMSS>` | Agent 名。显式命名 → 持久化；自动名 → 同样持久化，仅命名方便识别。 |
| `--prompt`、`-p` | string | 否 | `You are a helpful assistant.` | 系统提示词。 |
| `--model-service` | string | 否 | TTY 交互选 | ModelService 名；非 TTY 或 `--no-input` 时必填。 |
| `--model` | string | 否 | TTY 交互选 | ModelService 内的模型名。 |
| `--tool` | multi | 否 | | 工具名，可重复。 |
| `--skill` | multi | 否 | | 技能名，可重复。 |
| `--sandbox` | multi | 否 | | 沙箱名，可重复。 |
Expand All @@ -68,23 +66,21 @@ ar sa run [options]
| `--message`、`-m` | string | 否 | | 初始消息 —— 进入 REPL 后立刻发送。 |
| `--raw` | flag | 否 | false | 强制 raw SSE JSON 行输出。 |
| `--text-only` | flag | 否 | false | 只显示 Assistant 文本（隐藏工具调用）。 |
| `--no-input` | flag | 否 | false | 禁用交互式选择器；缺必填参数直接报错。 |
| `--no-input` | flag | 否 | false | 已弃用，行为为空操作；仅保留以兼容旧脚本。 |

### 示例

```bash
# 零配置 —— ModelService / 模型交互式选择
# 零配置 —— 服务端选用默认 model
ar sa run

# 指定 prompt + 初始消息
ar sa run -p "你是简洁风格的 Python 程序员" -m "写个 FizzBuzz"

# 非交互（脚本/CI）
# 启用工具
ar sa run \
--model-service svc-tongyi --model qwen-max \
--prompt "你是助手" \
--tool mcp-time-sa \
--no-input
--tool mcp-time-sa

# 命名让 Agent 持久保留
ar sa run --name my-helper -p "你是我的助手"
Expand Down Expand Up @@ -228,8 +224,6 @@ ar sa create --name <name> [options]
| `--name` | string | 是 | 全局唯一的 Agent 名。 |
| `--description` | string | 否 | 描述。 |
| `--prompt`、`-p` | string | 否 | 系统提示词。 |
| `--model-service` | string | 否 | ModelService 名。 |
| `--model` | string | 否 | 模型名。 |
| `--tool` | multi | 否 | 工具名，可重复。 |
| `--skill` | multi | 否 | 技能名，可重复。 |
| `--sandbox` | multi | 否 | 沙箱名，可重复。 |
Expand All @@ -240,12 +234,10 @@ ar sa create --name <name> [options]

```bash
ar sa create --name my-helper \
-p "你是我的助手" \
--model-service svc-tongyi --model qwen-max
-p "你是我的助手"

ar sa create --name researcher \
-p "深度调研助手" \
--model-service svc-tongyi --model qwen-max \
--tool web-search --tool mcp-time-sa \
--skill data-analyzer --skill report-generator
```
Expand Down Expand Up @@ -307,8 +299,6 @@ ar sa update <NAME> [options]
| `NAME` | 位置参数 | 要更新的 Agent。 |
| `--description` | string | 新描述。 |
| `--prompt`、`-p` | string | 新 prompt。 |
| `--model-service` | string | 新 ModelService。 |
| `--model` | string | 新模型名。 |
| `--tool` | multi | 替换 tools 列表。 |
| `--skill` | multi | 替换 skills 列表。 |
| `--sandbox` | multi | 替换 sandboxes 列表。 |
Expand All @@ -326,7 +316,6 @@ ar sa update <NAME> [options]

```bash
ar sa update my-helper -p "简洁风格的助手"
ar sa update my-helper --model-service svc-openai --model gpt-4o
ar sa update my-helper --tool web-search --tool calculator
ar sa update my-helper --clear-tools
```
Expand Down Expand Up @@ -399,9 +388,6 @@ metadata:
spec:
prompt: | # 可选
你是一个得力助手。
model: # 可选（但发起调用前必须有）
service: svc-tongyi
name: qwen-max
tools:
- mcp-time-sa
skills: []
Expand All @@ -417,8 +403,6 @@ spec:
| `metadata.name` | `name` |
| `metadata.description` | `description` |
| `spec.prompt` | `prompt` |
| `spec.model.service` | `model_service_name` |
| `spec.model.name` | `model_name` |
| `spec.tools` | `tools` |
| `spec.skills` | `skills` |
| `spec.sandboxes` | `sandboxes` |
Expand All @@ -436,15 +420,13 @@ metadata:
name: doc-writer
spec:
prompt: "你擅长撰写清晰的技术文档"
model: { service: svc-tongyi, name: qwen-max }
---
apiVersion: agentrun/v1
kind: SuperAgent
metadata:
name: code-reviewer
spec:
prompt: "你是资深 Python Reviewer"
model: { service: svc-tongyi, name: qwen-max }
```

`apply -f` 按文档顺序处理；任一失败会导致后续中断，**但已成功的 Agent 不会回滚**。
Expand Down
2 changes: 1 addition & 1 deletion pyproject.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,7 +20,7 @@ classifiers = [
]

dependencies = [
"agentrun-sdk[core]>=0.0.33",
"agentrun-sdk[core]>=0.0.34",
"pyyaml>=6.0",
"questionary>=2.0",
Copy link

Copilot AI Apr 25, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

questionary no longer appears to be used anywhere in the repo (the interactive model picker was removed, and there are no remaining imports/usages). Keeping it as a required runtime dependency increases install size and can introduce supply-chain/compatibility surface unnecessarily. Consider removing questionary>=2.0 from dependencies (or moving it to an optional extra if you plan to reintroduce interactive prompts elsewhere).

Suggested change
"questionary>=2.0",

Copilot uses AI. Check for mistakes.
]
Expand Down
Loading
Loading