WeWeRSS

订阅微信公众号等信源，获取原始文章内容用于整合日报

Security Scan

The skill's instructions, required tools, and environment access are consistent with its stated purpose (fetching RSS/articles from a WeWeRSS instance); no disproportionate credentials, installs, or privilege escalation are requested — though the SKILL.md requires a WEWERSS_BASE_URL which the registry metadata did not list.

Purpose & Capability

Name/description (subscribe to WeWeRSS/Open Channel to fetch original article content) match the runtime instructions. Required CLI tools (curl, jq) are appropriate for the described HTTP/JSON interactions. One minor metadata mismatch: the registry metadata listed no required env vars, but SKILL.md requires WEWERSS_BASE_URL — this is plausible (the service endpoint is necessary) but should be reconciled.

Instruction Scope

SKILL.md only instructs the agent to call the WeWeRSS Open Channel HTTP endpoints (create channel, add sources, list sources, fetch articles/feed). It does not instruct reading local files, other env vars, or sending data to unrelated external endpoints. Note: the instructions will send user-supplied URLs and channel/article data to the configured WEWERSS_BASE_URL, so any sensitive/internal URLs or content you provide could be transmitted to that server.

Install Mechanism

No install spec and no code files — instruction-only skill. This is low-risk from install perspective; nothing will be downloaded or written to disk by the skill itself.

Credentials

SKILL.md requires a single env var WEWERSS_BASE_URL (the service base URL), which is proportionate to the task. However, the registry metadata claimed 'required env vars: none' — reconcile this discrepancy. Also: ensure the base URL you provide points to a trusted instance because that server will receive content/URLs you submit.

Persistence & Privilege

Skill is not always-enabled and does not request elevated platform privileges. It does not attempt to modify other skills or system-wide settings.

Assessment

This skill appears coherent and low-risk if you point it at a trusted WeWeRSS instance. Before installing: (1) Confirm the SKILL.md-required env var WEWERSS_BASE_URL is documented in the registry and that you know which instance you'll use. (2) Only set WEWERSS_BASE_URL to a server you control or trust — that server will receive any source URLs and article content you submit, so do not send internal/private URLs to an untrusted endpoint. (3) Ensure curl and jq are available in the execution environment. (4) If you plan to use a public/shared WeWeRSS instance, review its privacy/storage policies because the skill does not request credentials but will transmit content to that service. The metadata mismatch about required env vars should be clarified by the publisher; resolving that would increase confidence.

WeWeRSS Open Channel Skill

通过 Open Channel API 订阅微信公众号等信源，获取原始文章数据。所有 AI 处理（摘要、分类、标签等）由 Agent 侧使用用户自己的 LLM token 完成。

Environment Variables

• WEWERSS_BASE_URL — WeWeRSS 实例地址（必填），例如 https://rss.example.com

创建频道

curl -s -X POST "$WEWERSS_BASE_URL/api/open/channels" \
  -H "Content-Type: application/json" \
  -d '{"name":"我的日报频道"}' | jq .

返回频道 UUID，后续操作均需要此 ID。

添加信源

提交任意 URL（公众号文章链接、RSS 地址、网站首页），系统自动发现 RSS feed：

CHANNEL_ID="your-channel-uuid"

# 添加 RSS feed
curl -s -X POST "$WEWERSS_BASE_URL/api/open/channels/$CHANNEL_ID/sources" \
  -H "Content-Type: application/json" \
  -d '{"url":"https://example.com/feed.xml"}' | jq .

# 添加网站（自动发现 RSS）
curl -s -X POST "$WEWERSS_BASE_URL/api/open/channels/$CHANNEL_ID/sources" \
  -H "Content-Type: application/json" \
  -d '{"url":"https://example.com"}' | jq .

查看信源列表

curl -s "$WEWERSS_BASE_URL/api/open/channels/$CHANNEL_ID/sources" | jq .

删除信源

SOURCE_ID="source-uuid-from-list"
curl -s -X DELETE "$WEWERSS_BASE_URL/api/open/channels/$CHANNEL_ID/sources/$SOURCE_ID"

获取文章

获取最近 N 天的原始文章内容（默认 1 天）：

# 最近 1 天
curl -s "$WEWERSS_BASE_URL/api/open/channels/$CHANNEL_ID/articles?days=1" | jq .

# 最近 7 天，第 2 页
curl -s "$WEWERSS_BASE_URL/api/open/channels/$CHANNEL_ID/articles?days=7&page=2&page_size=50" | jq .

返回字段仅包含原始数据：

• title - 文章标题
• author - 作者/公众号名
• link - 原文链接
• published_at - 发布时间
• description - RSS 自带摘要
• text_plain - 纯文本全文（可能为 null）
• source_name - 来源订阅名称

不返回任何 AI 生成字段（摘要、标签、分类、质量分等），Agent 需自行处理。

获取 Atom Feed

curl -s "$WEWERSS_BASE_URL/api/open/channels/$CHANNEL_ID/feed"

标准 Atom XML，可用于 RSS 阅读器或其他消费端。

速率限制

• 创建频道：10 次/小时（按 IP）
• 添加信源：30 次/小时（按频道）
• 文章/Feed 读取：120 次/小时（按频道）

典型使用流程

1. 创建频道 → 获得 UUID
2. 添加目标信源（公众号 RSS、行业博客等）
3. 每日定时获取文章 articles?days=1
4. Agent 用自己的 LLM 对文章进行摘要、分类、整合
5. 生成并推送日报