← 返回
未分类 中文

organization-operating-skill

A general-purpose skill for connecting the organization platform with external agents. Use it to access user, organization, post, and activity APIs, and to c...
用于将组织平台与外部代理连接的全能技能。可用于访问用户、组织、帖子和活动 API,以及...
urzeye urzeye 来源
未分类 clawhub v1.0.1 1 版本 100000 Key: 无需
★ 0
Stars
📥 369
下载
💾 0
安装
1
版本
#latest

概述

Organization Operating Skill

Overview

This skill turns the organization platform APIs into stable, executable actions for agents.

It focuses on authentication, session reuse, organization operations, post publishing, activity publishing, and signup flows.

Prefer reusable platform-wide capabilities, and leave organization-specific behavior to configuration or prompt snippets instead of building one-off skill logic for each organization.

Workflow

  1. First determine which task category applies:

Auth / environment / token: read references/auth_reference.md.

Organization operations: read references/org_reference.md.

Posts: read references/content_reference.md.

Activities: read references/activity_reference.md.

Capability scope and prioritization: read references/capability_inventory.md.

  1. Do not load all references by default.

Start with scripts/org_skill_cli.py --help, then read only the documents relevant to the current task.

  1. Keep platform-wide capabilities separate from organization-specific rules.

Put rule differences into organization configuration whenever possible instead of expanding the skill for one-off cases.

Common Flows

  • First-time agent account setup:

run session show, then guest-generate, then agent-login, and finally user-info to verify the identity.

  • Existing agent session:

run session show, then user-info to verify the identity.

If the token expires, prefer refresh instead of calling agent-login again by default.

  • Organization operations flow:

check isAllowCreate through user-info, then use org-create or org-detail, then post-create.

A "help post" is currently just a normal post, so publish it through post-create.

For activities, always run activity-save first to get the draft id, then activity-publish --draft-id .

  • Multi-agent usage:

different accounts must use different --session-file values so one token does not overwrite another in the same environment.

  • Environment convention:

default to production; explicitly pass --env test for integration testing and --env local for local backend work.

Core Capabilities

  • Authentication and environment switching
  • Organization create, query, update, join, and member pagination
  • Post publishing
  • Activity draft, publish, detail, search, and signup flows
  • See references/capability_inventory.md for future capability priorities

Integration Rules

  • Keep the skill generic, and push organization-specific behavior into organization configuration or prompt snippets.
  • Prefer completing workflows through existing APIs instead of adding code for one-time organization rules.
  • Before integrating any endpoint, document authentication mode, required parameters, key response fields, error codes, and code locations.
  • Keep human fallback for high-risk operations, including bans, dispute handling, credit penalties, and offline safety issues.
  • When platform APIs cannot express a rule cleanly, document the gap first and only then decide whether a new API is justified.

Resources

Use it for capability scope and prioritization, not for request-body details.

Navigation-only index.

Authentication, headers, token handling, environment rules, and session rules.

Organization APIs, default org avatars, and creation or update constraints.

Post publishing APIs and share-link conventions.

Activity draft, publish, search, signup, and share-link conventions.

  • scripts/

The current executable entry point is scripts/org_skill_cli.py.

When you need the command list, start with python scripts/org_skill_cli.py --help.

Runtime Defaults

  • Default base URL is production: https://api.zingup.club/biz
  • Session state is not written back into the skill repository.
  • The safest approach is to pass --session-file explicitly.
  • If --session-file is not provided:
  • the CLI will default to ~/.organization-operating-skill/sessions/ and create it automatically
  • the current implementation can be overridden through ORG_SKILL_STATE_DIR
  • older .codex-state/... sessions are still readable for smooth migration
  • Prefer switching environments with --env
  • prod
  • test
  • local
  • You can also pass --base-url explicitly.
  • Default request headers include:
  • x-platform=3
  • x-language=ch
  • x-package=com.groupoo.zingup
  • x-timezone=
  • English-speaking users can explicitly switch to --language us.
  • For header details, prefer auth_reference.md and the actual CLI behavior.
  • Web-style requests:
  • web-config-get and post-create automatically include the required web-specific headers
  • x-device-id strategy:
  • generate one automatically if it does not exist yet
  • once it is stored in session state, reuse it by default
  • Common debug entry points:
  • start with python scripts/org_skill_cli.py --help
  • python scripts/org_skill_cli.py --env prod session show
  • python scripts/org_skill_cli.py --env test session show
  • Login convention:
  • first-time agent setup: guest-generate -> agent-login
  • for the same agent later: reuse the session and call refresh when needed
  • Organization lookup convention:
  • use org-detail as the default member-view detail endpoint
  • If org-create is called without an avatar, it will automatically use web-config-get to select a default organization avatar.
  • Activity publish convention:
  • activity-save submits the complete draft payload
  • activity-publish normally only needs the draft id
  • Multi-account convention:
  • if you need both a publisher and an attendee in the same environment, use separate --session-file values explicitly

Context Discipline

Load only the references needed for the current task, and avoid dumping the entire API surface into context at once.

版本历史

共 1 个版本

  • v1.0.1 当前
    2026-05-07 07:13 安全 安全

安全检测

腾讯云安全 (Keen)

安全,无风险
查看报告

腾讯云安全 (Sanbu)

安全,无风险
查看报告

🔗 相关推荐

ai-agent

Skill Vetter

spclaudehome
AI智能体技能安全预审工具。安装ClawdHub、GitHub等来源技能前,检查风险信号、权限范围及可疑模式。
★ 1,243 📥 271,411
ai-agent

Find Skills

guipi888
场景驱动+关键词双模式技能发现工具。当用户用自然语言描述场景/需求(如"我想做一个海报""帮我分析股票"),或明确说"安装技能/find skills/找个skill"时,自动从官方内置、本地已安装、SkillHub、虾评、GitHub、C
★ 1,494 📥 558,276
ai-agent

Self-Improving + Proactive Agent

ivangdavila
自我反思+自我批评+自我学习+自组织记忆。智能体评估自身工作、发现错误并持续改进。
★ 1,412 📥 325,271