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 OmerLapidot/whatsapp-connector --skill whatsappgit clone --depth 1 https://github.com/OmerLapidot/whatsapp-connectorWrote 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/omerlapidot/whatsapp-connector/whatsapp)<a href="https://agentmods.dev/skills/omerlapidot/whatsapp-connector/whatsapp"><img src="https://agentmods.dev/badge/skills/omerlapidot/whatsapp-connector/whatsapp/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/omerlapidot/whatsapp-connector/whatsapp"><img src="https://agentmods.dev/badge/skills/omerlapidot/whatsapp-connector/whatsapp.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 | $0.00066 | $0.02019 |
| Opus 5 | $0.00033 | $0.01009 |
| Sonnet 5 | $0.00013 | $0.00404 |
| Haiku 4.5 | $0.00007 | $0.00202 |
Grade A, and why
whatsapp 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 — 131 lines — stays where its author put it; the contents beside it link to each section on GitHub.
WhatsApp connector
Drive the user's WhatsApp via the wa CLI, which talks to an always-on local daemon.
Always call it by absolute path: {{WA_DIR}}/bin/wa
Reading (allowed freely)
wa chats— list chats/groupswa read "<chat>" [--limit 30 | --all]— recent messages (each has anidfor reactions).--allreads the WHOLE chat (every synced message), bypassing the 200 cap — use sparingly: a busy chat can be thousands of messages (large payload + heavy on your context)wa search "<query>" [--chat "<chat>"]— find messageswa members "<group>"— who is in a groupwa contacts "<query>"— resolve a name/numberwa media "<chat>" "<message-id>"— download an image/file; prints a path you can then open
Chats are addressed by name (fuzzy) or exact id. If wa reports the name is ambiguous, show the candidates and ask which one.
Response shapes (verified against a live account)
Everything prints as JSON (or a plain string). Shapes you can rely on:
wa chats→ array of{ id, name, unread }, ordered most-recently-active first. Theidsuffix is the chat type:…@g.us— a group…@lid— an individual person (WhatsApp's current per-contact id). The older…@c.usform can also appear; treat BOTH as 1:1 people. To filter to real people, exclude@g.us— do NOT match on@c.usalone (modern accounts return@lidand you'll get zero results).
wa read→{ chat, id, messages: [ … ] }, oldest→newest. Each message is{ id, sender, ts, text, hasMedia }:senderis the literal string"me"for the user's own messages, otherwise the counterparty's chat id — so message direction is justsender === "me".tsis a UNIX timestamp in seconds (multiply by 1000 for a JSDate).--limitis capped at 200 (MAX_LIMIT); a busier chat returns only its most recent 200, so a chat sitting at exactly 200 means "≥200", not exactly 200. Pass--allto bypass the cap and fetch every message — but "every" means every message currently synced into the Web session, NOT the full lifetime history. Depth is volatile: a freshly linked device backfills older messages over time, and a daemon restart flushes the loaded depth — right after a (re)start most chats read only their newest message or two and re-deepen as the session re-syncs. So a low--allcount can mean "not synced yet," not "that's the whole chat." A big, warmed chat can return thousands of messages at once.
wa status→{ state, ready, syncPercent }(plusrecovery: {attempt, max}while self-recovering);state∈starting | needs-login | syncing | ready | recovering | relinking | auth-failure | disconnected.- Errors come back as
{ ok:false, error, code }with codes likeNOT_ALLOWED(chat not on the allow-list),NOT_APPROVED(human denied or the dialog timed out),AMBIGUOUS(name matched >1 chat),NOT_FOUND.
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 · 131 lines · 66 tokens per session scan A f0acb15393d2
whatsapp is a skill published in the GitHub repository OmerLapidot/whatsapp-connector (6 stars, last pushed 1mo ago), licensed MIT. It adds 66 tokens to every session and 2,019 once invoked, about $0.0003 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 2026-08-31.
Other skills, from other repositories
orbit-notion
Open Orbit briefing skill — selected by the Orbit pipeline when Notion is the user's only connected connector, or when the user explicitly scopes their daily digest to Notion. Pulls the past 24 hours of document edits, comments, mentions, and database row changes from the user's authenticated Notion connection and…
Cortex
Operate Cortex, the LifeOS memory system — the typed Knowledge Archive (People, Companies, Ideas, Research with typed related: links) plus recall of prior work sessions, ISAs, and conversations. Search, add, harvest, develop, ingest, distill, graph-navigate, recall. USE WHEN cortex, knowledge, knowledge base, search…
pinchtab-mcp
Use this skill when a task requires browser automation through PinchTab's MCP server connected to a remote browser instance. Covers navigation, element interaction, data extraction, form filling, multi-step flows, and session management via MCP tools.
feishu
Work with Feishu or Lark bots, docs, sheets, bitables, approval flows, and OpenAPI/MCP setup without hardcoding credentials.
peekaboo
Capture and automate macOS UI with the Peekaboo CLI.
mochi-remind
Handle due reminders — notify the user with natural language and mark them done.