Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add 1xiaoyueryuer/boss-hr-agent-toolkit --skill sharedgit clone --depth 1 https://github.com/1xiaoyueryuer/boss-hr-agent-toolkitWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/1xiaoyueryuer/boss-hr-agent-toolkit/shared)<a href="https://agentmods.dev/skills/1xiaoyueryuer/boss-hr-agent-toolkit/shared"><img src="https://agentmods.dev/badge/skills/1xiaoyueryuer/boss-hr-agent-toolkit/shared/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/1xiaoyueryuer/boss-hr-agent-toolkit/shared"><img src="https://agentmods.dev/badge/skills/1xiaoyueryuer/boss-hr-agent-toolkit/shared.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00260 | $0.03474 |
| Opus 5 | $0.00130 | $0.01737 |
| Sonnet 5 | $0.00052 | $0.00695 |
| Haiku 4.5 | $0.00026 | $0.00347 |
Grade A, and why
shared scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 288 lines — stays where its author put it; the contents beside it link to each section on GitHub.
shared —— 基础模块
本目录是全工具包依赖的 Python 库。不是入口，不被任何 AI 智能体当作 Skill 直接加载。 业务脚本（
boss_jd.py/recommend_list.py等）import这些模块即可。
模块清单
| 模块 | 角色 | 主要导出 |
|---|---|---|
cdp_preflight.py |
CDP 连接 + 登录态探测 | connect_cdp, CDPSession, check_login, get_cookies, RECRUITER_REQUIRED_COOKIES |
recruiter_job_catalog.py |
招聘者岗位目录 | list_jobs, resolve_recruiter_job, fetch_job_detail, JOB_LIST_URL, JOB_EDIT_URL |
output_manager.py |
文件路径管理（JobOutputManager） |
JobOutputManager, resolve_encrypt_job_id, OUTPUT_ROOT |
run_orchestrator.py |
run_id 生命周期（RunOrchestrator） |
create_new_run, bind_existing_run, finish |
job_resume_store.py |
跨 run 累计简历/评分去重 | JobResumeStore, is_scored, mark_scored |
job_registry.py |
encryptJobId ↔ jobName 元数据 | JobRegistry |
cli_runner.py |
Windows PowerShell 安全的 CLI 执行层 | run_python_cli（白名单 9 个 tool） |
confirm_run.py |
用户确认门（翻 run.json.confirmed=true） |
CLI 子命令 |
fix_encoding.py |
Windows UTF-8 stdout | 自动 import 即生效 |
human_interaction.py |
智能体交互辅助 | — |
cdp_preflight.py —— CDP 连接 + 登录态探测
设计动机
本工具包不依赖任何第三方 CLI（包括早期项目曾用的 boss_agent_cli
的 boss login / boss me / boss status 等登录检测命令）。本模块用 patchright
直连 Edge CDP，自己判断：
- Edge 是否在指定端口（如
http://localhost:9222）跑着 - BOSS 招聘者 session 是否有效（
zp_at+wt2+bst三 cookie 都在且非空） - 当前页面是
recommend/chat/job_edit/login/unknown
所有 BOSS HTTP 调用（recruiter_job_catalog 的 list_jobs 等）都走浏览器内 fetch（page.evaluate('fetch(...)')），
复用浏览器真实 TLS 指纹 + 自动带 cookie，不需要单独同步 __zp_stoken__。
接口
from cdp_preflight import (
connect_cdp, # 函数
check_login, # 函数
get_cookies, # 函数
CDPSession, # dataclass
RECRUITER_REQUIRED_COOKIES, # tuple: ('zp_at', 'wt2', 'bst')
)
# 1) 连 Edge
session = connect_cdp() # 默认 http://localhost:9222
session = connect_cdp("http://1.2.3.4:9222", timeout_ms=15000)
# 2) 检查登录
state = check_login(session)
# state = {
# 'logged_in': True/False,
# 'cookies': {
# 'zp_at': True/False,
# 'wt2': True/False,
# 'bst': True/False,
# 'present': [...], # 所有 cookie 名
# 'total': int,
# },
# 'current_url': 'https://www.zhipin.com/web/chat/recommend',
# 'page_kind': 'recommend' | 'chat' | 'job_edit' | 'login' | 'unknown',
# }
# 3) 取 cookie（HTTP 库复用）
cookies = get_cookies(session) # {'zp_at': '...', 'wt2': '...', ...}
# 4) 关闭（不关 Edge，只关 playwright 进程）
session.disconnect()
What ships with it
10 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- cdp_preflight.py 5.6 KB runs code
- cli_runner.py 11 KB runs code
- confirm_run.py 4.3 KB runs code
- fix_encoding.py 1.3 KB runs code
- human_interaction.py 4.9 KB runs code
- job_registry.py 4.2 KB runs code
- job_resume_store.py 15 KB runs code
- output_manager.py 15 KB runs code
- recruiter_job_catalog.py 11 KB runs code
- run_orchestrator.py 10 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 288 lines · 260 tokens per session scan A efc73e46196e
shared is a skill published in the GitHub repository 1xiaoyueryuer/boss-hr-agent-toolkit (46 stars, last pushed 1mo ago), licensed MIT. It adds 260 tokens to every session and 3,474 once invoked, about $0.0013 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
qiaomu-opencli-explorer
A guide for creating OpenCLI adapters for websites. An adapter is code that lets a command-line tool communicate with a particular website’s interfaces and data.
firecrawl-build-scrape
Integrate Firecrawl /scrape into product code for single-page extraction. Use when an app already has a URL and needs markdown, HTML, links, screenshots, metadata, or structured page output. Prefer this skill over broader crawl patterns when the feature is page-level.
gakido
High-performance Python HTTP client with browser impersonation and anti-bot evasion. Use when the user needs to bypass bot detection, impersonate browsers with TLS fingerprinting, make HTTP/2/3 requests, or build web scrapers that evade anti-bot systems.
derive-client
Reverse-engineer a website's internal API by recording browser traffic into a HAR file, then generate a standalone client or CLI that calls the endpoints directly, with no browser needed after the first recording. Use when asked to "derive a client", "build a CLI for ", "reverse engineer this site's API", "record…
computer-use
OS/window-level inspection and input in visible local app windows through orca computer: native apps, external browser windows (Chrome, Edge, Safari), and app webviews. Not for Orca's embedded browser (use orca-cli) or page-only automation (use Playwright or CDP).
opencli-sitemap-author
Use when creating or maintaining OpenCLI site sitemaps: agent-facing navigation, page-state, action, workflow, API-reference, pitfall, and fallback knowledge for a website. Use after browser exploration discovers durable site context, when a sitemap is stale, or when promoting local site knowledge into the repo.