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 Haochenhust/ch-skills --skill wechat-feedsgit clone --depth 1 https://github.com/Haochenhust/ch-skillsWrote 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/haochenhust/ch-skills/wechat-feeds)<a href="https://agentmods.dev/skills/haochenhust/ch-skills/wechat-feeds"><img src="https://agentmods.dev/badge/skills/haochenhust/ch-skills/wechat-feeds/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/haochenhust/ch-skills/wechat-feeds"><img src="https://agentmods.dev/badge/skills/haochenhust/ch-skills/wechat-feeds.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00105 | $0.02230 |
| Opus 5 | $0.00053 | $0.01115 |
| Sonnet 5 | $0.00021 | $0.00446 |
| Haiku 4.5 | $0.00011 | $0.00223 |
Grade A, and why
wechat-feeds scanned grade A with 1 finding 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 9d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- CDP proxy healthy: `curl -s http://localhost:3456/health` → expect `"connected":true`. If no response, the proxy isn't running — start it (`node cdp-proxy.mjs`) and make sure Chrome is up with remote debugging. How it starts
The opening of the file, as written. The whole thing — 150 lines — stays where its author put it; the contents beside it link to each section on GitHub.
wechat-feeds
Build and maintain a local library of WeChat Official Account (微信公众号) articles. The bundled tool (tool/) logs into the 公众号 admin platform in your own Chrome, then pulls each subscribed account's article list and full text into data/<account>/<date>-<title>.md, with a SQLite index (feeds.db) for querying.
Two things this skill does:
- Set up / maintain the library — install, log in, subscribe to accounts.
- Incrementally update — pull new articles up to now (the
update-feedsworkflow below), including automatic session re-login handling.
How it works (architecture)
CLI (bun, tool/src) ──HTTP──▶ cdp-proxy.mjs ──WebSocket(CDP)──▶ your Chrome (logged into mp.weixin.qq.com)
- The scraper never talks to WeChat directly. It asks the CDP proxy to run
fetch()inside a logged-in mp.weixin.qq.com tab, so every request is same-origin and carries the real login cookie — no separate auth, no easy anti-crawl trip. - Login state is a
token(insession.json) plus the browser's cookie. The token expires often; the cookie lasts a few days.refreshre-derives the token from a still-valid cookie; when the cookie itself dies you re-login in the browser once.
Prerequisites
- Bun (runs the CLI; uses
bun:sqlite) — https://bun.sh - Node.js 22+ or Bun (runs the CDP proxy; needs a global
WebSocket) - Google Chrome you can start with remote debugging
- A WeChat Official Account you can log into
mp.weixin.qq.comwith (any account works — the 图文素材 search reads other public accounts' published articles)
One-time setup
All commands run from the tool/ directory.
cd tool
bun install
# 1. Start Chrome with remote debugging (keep it running).
# macOS:
/Applications/Google\ Chrome.app/Contents/MacOS/Google\ Chrome --remote-debugging-port=9222
# Linux: google-chrome --remote-debugging-port=9222
# Windows: chrome.exe --remote-debugging-port=9222
# 2. Start the CDP proxy (separate terminal; keep it running).
node cdp-proxy.mjs # or: bun cdp-proxy.mjs
# 3. Log in — opens the 公众号 login page; scan with WeChat.
bun run src/cli.ts login
# 4. Subscribe to accounts (by name or by a share link to one of their articles).
bun run src/cli.ts add 公众号名称
bun run src/cli.ts add https://mp.weixin.qq.com/s/xxxxx
bun run src/cli.ts list
What ships with it
18 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.
- tool/.gitignore 492 B
- tool/cdp-proxy.mjs 18 KB runs code
- tool/config.example.json 131 B
- tool/package.json 321 B
- tool/src/api/mp.ts 7.0 KB runs code
- tool/src/api/request.ts 3.7 KB runs code
- tool/src/auth/login.ts 1.9 KB runs code
- tool/src/auth/refresh.ts 2.5 KB runs code
- tool/src/auth/relogin.ts 1.2 KB runs code
- tool/src/auth/session.ts 499 B runs code
- tool/src/cli.ts 6.2 KB runs code
- tool/src/config.ts 2.0 KB runs code
- tool/src/converter/html2md.ts 2.3 KB runs code
- tool/src/converter/images.ts 3.1 KB runs code
- tool/src/fetcher/pull.ts 8.4 KB runs code
- tool/src/notify/notify.ts 1.2 KB runs code
- tool/src/store/db.ts 2.3 KB runs code
- tool/tsconfig.json 327 B
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.
- 9d ago First seen · 150 lines · 105 tokens per session scan A 4e0fa3d03ae0
wechat-feeds is a skill published in the GitHub repository Haochenhust/ch-skills (9 stars, last pushed 1mo ago), licensed MIT. It adds 105 tokens to every session and 2,230 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
storyboard-designer
A tool for turning a narration or story script into a structured storyboard spreadsheet for AI video projects. The spreadsheet includes shot planning, character design, first- and last-frame prompts, narration, and sound effects.
pydicom
Use pydicom to read, inspect, write, transform, and safely preflight local DICOM datasets and pixel data. Applies to DICOM metadata, transfer syntaxes, compression plugins, frames, private elements, JSON, and bounded de-identification review.
venue-templates
Prepare journal manuscripts, conference papers, research posters, and grant documents using venue-specific formatting guidance and bundled LaTeX scaffolds. Use when selecting an official template, checking current page or anonymity rules, adapting academic writing to a venue, or inspecting a submission PDF.
treatment-plans
Format and structurally validate local treatment-plan documentation after clinical decisions have already been supplied and verified by authorized licensed professionals. Use for source traceability, clinician-authored intervention records, goals and checkpoints, shared-decision records, reconciliation handoffs, and…
xlsx
Create, edit, analyze, or convert Excel spreadsheets (.xlsx, .xlsm, .xltx) where the workbook file is the primary deliverable. Use for formulas, formatting, financial models, multi-sheet workbooks, and tabular cleanup exported to Excel. Also applies to .csv/.tsv when the user wants spreadsheet output. Do NOT use for…
baoyu-youtube-transcript
A tool for downloading the written captions, subtitles, chapter information, speaker labels, and cover image from a YouTube video using its URL or ID.