mirror of
https://git.openapi.site/https://github.com/desirecore/market.git
synced 2026-10-21 09:24:09 +08:00
feat: 创建与更新技能支持独立长期指令 (#211)
创建与更新技能增加可选 instructions 字符串：普通 Markdown 全文、省略保留、字符串整体替换、空字符串清空。内容归属明确为人格、原则、长期指令与技能方法；既有 AgentFS 直接编辑遵循实际权限与审批。create-agent 2.8.0 / update-agent 3.4.0 的最低客户端为 10.0.179，双语正文与 sidecar 同步。 Add optional plain-Markdown instructions to Agent creation and update guidance, including preserve/replace/clear semantics and clear content ownership. Existing direct AgentFS editing retains its permission and approval rules. Both localized bodies and catalog sidecars require DesireCore 10.0.179. 验证 / Validation：目录与 i18n 校验 0 errors；严格 sidecar 全覆盖检查 0 errors（既有证据警告保留）；双语由人工同步并按真实 source body 计算 freshness hash。依据用户明确授权，本次以全树凭据形状扫描、Git/PR 元数据扫描与客户中立性人工审阅代替缺失的私有敏感词表检查；凭据形状扫描 740 文件无命中，不声称执行私有词表匹配。
This commit is contained in:
GitHub
@@ -2,8 +2,8 @@
|
||||
license: "MIT"
|
||||
name: create-agent
|
||||
description: >-
|
||||
通过多轮对话收集需求，调用 ManageAgent 内置工具创建新的 AgentFS v2 智能体，支持自定义 persona 和 principles。Use when 用户要求创建新智能体、培养某领域助手、或快速基于模板生成可治理 Agent。可选头像生成需配置图像服务，适用供应商条款/额度及独立费用。
|
||||
version: 2.7.0
|
||||
通过多轮对话收集需求，调用 ManageAgent 内置工具创建新的 AgentFS v2 智能体，支持自定义 persona、principles 和可选 instructions。Use when 用户要求创建新智能体、培养某领域助手、或快速基于模板生成可治理 Agent。可选头像生成需配置图像服务，适用供应商条款/额度及独立费用。
|
||||
version: 2.8.0
|
||||
compatibility: "DesireCore ManageAgent; optional avatar generation requires a configured image provider under its own terms and quota, with separate charges; unavailable generation retains the existing avatar."
|
||||
type: meta
|
||||
risk_level: low
|
||||
@@ -15,7 +15,7 @@ tags:
|
||||
- meta
|
||||
metadata:
|
||||
author: desirecore
|
||||
updated_at: '2026-10-03'
|
||||
updated_at: '2026-10-08'
|
||||
i18n:
|
||||
default_locale: en-US
|
||||
source_locale: zh-CN
|
||||
@@ -26,17 +26,17 @@ metadata:
|
||||
name: 创建智能体
|
||||
short_desc: 通过自然语言对话收集需求，一键创建专业化数字智能体
|
||||
description: >-
|
||||
通过多轮对话收集需求，调用 ManageAgent 内置工具创建新的 AgentFS v2 智能体，支持自定义 persona 和 principles。Use when 用户要求创建新智能体、培养某领域助手、或快速基于模板生成可治理 Agent。可选头像生成需配置图像服务，适用供应商条款/额度及独立费用。
|
||||
通过多轮对话收集需求，调用 ManageAgent 内置工具创建新的 AgentFS v2 智能体，支持自定义 persona、principles 和可选 instructions。Use when 用户要求创建新智能体、培养某领域助手、或快速基于模板生成可治理 Agent。可选头像生成需配置图像服务，适用供应商条款/额度及独立费用。
|
||||
body: ./SKILL.zh-CN.md
|
||||
source_hash: sha256:7ffb7a362c26a91e
|
||||
source_hash: sha256:3c0613f2d9e9479a
|
||||
translated_by: human
|
||||
en-US:
|
||||
name: Create Agent
|
||||
short_desc: Collect requirements through natural-language conversation and create a specialized digital Agent in one step
|
||||
description: >-
|
||||
Collect requirements through multi-turn conversation and call the ManageAgent builtin tool to create a new AgentFS v2 Agent, with customizable persona and principles. Use when the user asks to create a new Agent, raise a domain assistant, or quickly produce a governable Agent from a template. Optional avatar generation requires a configured image service under provider terms/quota and separate charges.
|
||||
Collect requirements through multi-turn conversation and call the ManageAgent builtin tool to create a new AgentFS v2 Agent, with customizable persona, principles and optional standing instructions. Use when the user asks to create a new Agent, raise a domain assistant, or quickly produce a governable Agent from a template. Optional avatar generation requires a configured image service under provider terms/quota and separate charges.
|
||||
body: ./SKILL.md
|
||||
source_hash: sha256:7ffb7a362c26a91e
|
||||
source_hash: sha256:3c0613f2d9e9479a
|
||||
translated_by: human
|
||||
market:
|
||||
icon: >-
|
||||
@@ -58,7 +58,7 @@ market:
|
||||
verified: true
|
||||
compatible_agents: []
|
||||
channel: latest
|
||||
required_client_version: 10.0.108
|
||||
required_client_version: 10.0.179
|
||||
---
|
||||
|
||||
# create-agent skill
|
||||
@@ -69,7 +69,7 @@ Collect requirements through natural-language conversation and call the ManageAg
|
||||
|
||||
## L1: Overview
|
||||
|
||||
Meta-skill: gather requirements over multi-turn conversation → generate persona/principles → land via `ManageAgent(action='create')`. Use it to raise a domain specialist (legal advisor, financial analyst), deploy a customized business Agent quickly, or produce a prototype from a template. The created repo conforms to the AgentFS v2 spec and is version-managed by git (governable, traceable). Its value is what the tool can't give: domain-tailored persona/principles generation + a pre-create preview confirmation.
|
||||
Meta-skill: gather requirements over multi-turn conversation → generate persona/principles and standing instructions → land via `ManageAgent(action='create')`. Use it to raise a domain specialist (legal advisor, financial analyst), deploy a customized business Agent quickly, or produce a prototype from a template. The created repo conforms to the AgentFS v2 spec and is version-managed by git (governable, traceable). Its value is what the tool can't give: domain-tailored persona/principles generation + a pre-create preview confirmation.
|
||||
|
||||
## L2: Detailed Spec
|
||||
|
||||
@@ -87,9 +87,10 @@ Trigger (any): user explicitly says "create an Agent / make me an assistant"; de
|
||||
|
||||
### Stage 3: Content Generation
|
||||
|
||||
Organize persona and principles as structured fields (do not emit raw markdown; present field-by-field). Leave uncollected fields empty for the system to fill with defaults:
|
||||
Organize persona, principles and optional standing instructions as structured fields (do not emit raw markdown; present field-by-field). Leave uncollected fields empty for the system to fill with defaults:
|
||||
|
||||
- **persona**: L0 one-sentence core identity; L1 `role` / `personality` (array of trait tags) / `communication_style`; L2 specialty, values, decision preferences (free-form).
|
||||
- **instructions**: optional plain Markdown string for stable responsibilities, default work strategies and delivery requirements; the full text is loaded without L0/L1/L2 parsing or truncation. Detailed procedures and reference material belong to skills. If only the name is known, omit this file.
|
||||
- **principles**: L0 one-sentence top principle; L1 `must_do` / `must_not` (safety red lines) / `priority`; L2 governance principles, escalation rules (free-form).
|
||||
|
||||
**Domain matching reference** (recommend personality and red lines by domain):
|
||||
@@ -103,7 +104,7 @@ Organize persona and principles as structured fields (do not emit raw markdown;
|
||||
|
||||
### Stage 4: User Confirmation
|
||||
|
||||
Present the preview in natural language / tables (name, description, persona, principles; **no raw markdown source**), e.g.:
|
||||
Present the preview in natural language / tables (name, description, persona, principles, and optional standing instructions; **no raw markdown source**), e.g.:
|
||||
|
||||
> About to create "Legal Advisor Assistant" — focused on contract review and legal risk assessment.
|
||||
> **Persona**: digital legal advisor; professional, rigorous, prudent; uses legal terms accurately with plain-language explanations.
|
||||
@@ -122,7 +123,8 @@ ManageAgent({
|
||||
"name": "Legal Advisor Assistant",
|
||||
"description": "A digital Agent focused on contract review and legal risk assessment",
|
||||
"persona": { "L0": "…", "L1": { "role": "…", "personality": ["professional","rigorous","prudent"], "communication_style": "…" } },
|
||||
"principles": { "L0": "…", "L1": { "must_do": ["…"], "must_not": ["…"], "priority": "user safety > accuracy > efficiency" } }
|
||||
"principles": { "L0": "…", "L1": { "must_do": ["…"], "must_not": ["…"], "priority": "user safety > accuracy > efficiency" } },
|
||||
"instructions": "# Responsibilities\nDescribe stable duties here.\n\n## Delivery requirements\nInclude the result location and evidence."
|
||||
})
|
||||
```
|
||||
|
||||
@@ -189,3 +191,5 @@ Present in a user-friendly way (no internal paths / technical details): confirm
|
||||
|
||||
- AgentFS structure, troubleshooting, and protected paths: see `_agentfs-background.md` and `_protected-paths.yaml`.
|
||||
- Always create via `ManageAgent`; never use curl / HTTP or write AgentFS directories directly. Requires client ≥ 10.0.90.
|
||||
|
||||
Keep role/expression in persona, boundaries in principles, stable duties in instructions, and methods in skills. Never place private user or team data in distributable Agent files.
|
||||
|
||||
Reference in New Issue
Block a user