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 philipyaz/cos --skill whatsapp-triagegit clone --depth 1 https://github.com/philipyaz/cosWrote 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/philipyaz/cos/whatsapp-triage)<a href="https://agentmods.dev/skills/philipyaz/cos/whatsapp-triage"><img src="https://agentmods.dev/badge/skills/philipyaz/cos/whatsapp-triage/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/philipyaz/cos/whatsapp-triage"><img src="https://agentmods.dev/badge/skills/philipyaz/cos/whatsapp-triage.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 2 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Excessive Agency · line 58 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
- medium Excessive Agency · line 65 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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 | $0.00144 | $0.16928 |
| Opus 5 | $0.00072 | $0.08464 |
| Sonnet 5 | $0.00029 | $0.03386 |
| Haiku 4.5 | $0.00014 | $0.01693 |
Grade A, and why
whatsapp-triage 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 8d 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.
4.5)** — never `bash`/`curl` (Cowork's sandbox blocks outbound HTTP; the tools exist How it starts
The opening of the file, as written. The whole thing — 894 lines — stays where its author put it; the contents beside it link to each section on GitHub.
WhatsApp → Board (the WhatsApp reconciler)
This skill reconciles WhatsApp with the board. For every chat with recent
activity — DMs and groups, inbound and the user's own sent messages — it links
the message onto the right case, closes or adds tasks, moves the lane, sets labels,
puts any confirmed appointment on the board calendar (Step 4), and keeps one
card per matter. It writes only through the chief-of-staff MCP tools — the
board MCP for cases / tasks / reminders / messages / labels, the
calendar MCP for events, and the nutrition MCP for shopping items (STEP
4.5) — never bash/curl (Cowork's sandbox blocks outbound HTTP; the tools exist
for exactly this).
This skill is BOARD-ONLY. It uses only the READ tools of the whatsapp
MCP — search_contacts, get_contact, list_chats, get_chat, list_messages,
get_message_context, get_last_interaction, get_direct_chat_by_contact,
get_contact_chats, download_media. It never calls send_message,
send_file, or send_audio_message; it never sends or drafts a WhatsApp message,
ever. It owns the board side only: the knowledge in a message (a fact, a
decision, new context about a sender) is submitted to the vault through the vault
MCP's async ingest, driven per /vault-operations; the vault synthesizes
knowledge, this skill reconciles cases. (The vault is knowledge-only — never write
task checkboxes into wiki pages; open work lives on the board.)
The headline guardrail. The board is a shared surface: the human edits it by hand in the UI, the agent edits it via this skill. A message can make the agent think a matter is open when the user has already closed it by hand. The agent must never undo a manual action. STEP 3 is the contract; read it before any write.
The JID gotcha — keep it visible the whole sweep. WhatsApp identifies a chat in three forms: a DM is
<phone>@s.whatsapp.net, a group is<id>@g.us, and an anonymous link-id is<id>@lid. The same person can appear as both a@s.whatsapp.netphone and a@lid— most "this person has two cards" bugs trace back to not collapsing those two forms (Step 7). Always think about which form you're holding.
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.
- 8d ago Changed · +114 lines 036946f58a81
- 12d ago First seen · 780 lines · 144 tokens per session scan A f2cb86e3ea94
whatsapp-triage is a skill published in the GitHub repository philipyaz/cos (4 stars, last pushed 2d ago), licensed MIT. It adds 144 tokens to every session and 16,928 once invoked, about $0.0007 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
wiki-fold
Roll up the wiki's journal.md entries into structured fold pages — like 2^k log compaction. Reads the last 2, 4, 8, 16... entries and writes a fold page that summarizes them by extractive summarization (no invention), with backlinks to children. Idempotent at the structural level — re-running with the same window…
decision-consolidate
Consolidate a SETTLED decision page (accepted / superseded / rejected — never proposed): compress the body to canon (verdict byte-intact, minimal why, alternatives as a table) and move the deliberation chronicle to a verified archives/ note (type: decision-archive) excluded from recall and search. Compress + archive…
fizzy-tasks
Fetch open Fizzy cards assigned to the user, open bugs, and unassigned tasks. Presents a prioritised digest with deep links. Use when user says "fizzy tasks", "my fizzy cards", "what are my open tasks in fizzy", "fizzy action items", "bugs in fizzy", "show my tasks", or wants a digest of current Fizzy workload.
fizzy-write
Create cards, create boards, update cards, close/reopen cards, add comments, assign users, or move cards to a column in Fizzy. Use when user says "create card", "create board", "add task to fizzy", "close fizzy card", "comment on fizzy card", "assign fizzy card", "move card to column", or anything involving writing to…
fizzy-board-monitor
Show an on-demand Kanban overview of a Fizzy board — cards grouped by column with counts. Default board is the QA board. Use when user says "show fizzy board", "fizzy board overview", "list fizzy cards", "what's on the board", "show [board name] board", or wants to see the current state of any Fizzy board.
fizzy-product-manager
Product management knowledge for Fizzy Kanban boards — Kanban principles, API workflows, triage, hygiene, and reporting.