OR Key
drop another .md file to compare - side-by-side diff against whatsapp

whatsapp

Reads and sends your WhatsApp messages.
description: "Triggers on prompt mention of 'whatsapp'."
personal 2 files 10 recent evals

What it does for you

Reads and sends your WhatsApp messages.

What it produces

A recent result, so you can see the kind of work it returns.

loading…

How to get it

These run inside the Snappy workspace. Want this working in your business? I set skills like this up with you, in one focused week.

Work with me
For developers how this skill is built, graded, and how it runs

at a glance- the short version

actorExported functions in state/lib/whatsapp.ts.
auditorNone wired yet - eval is manual (Robert review).
eval modeshape
categoryChannels
stages1
dependssettings

what's inside - the parts that make up a skill 3/4 present

A skill is just a few plain-text files. Only the main one is required. The rest are optional, added as the work needs them. This is what the skill is made of; how it runs is just below.

The skill
state/skills/whatsapp/SKILL.md present
the skill itself, in plain text
The main file. It says what the skill is and lays out the steps in plain English.
Code
state/lib/whatsapp.ts present
code the skill can run
Reusable code this skill can call when it needs to.
Scripts
state/bin/whatsapp/ not present
helper scripts
Optional. Added when a skill has a few commands to run.
Loader
state/skills/whatsapp/AGENTS.md present
what the AI loads on the fly
Loaded automatically the moment this skill is needed. Kept short on purpose.

how it's graded - what counts as a good run 4 criteria · 2 deterministic · 2 judge

Each row is one thing a good run has to get right. deterministic means a quick check decides, pass or fail. judge means the AI reads the result and rates it. Grading each piece on its own (instead of one overall score) shows exactly where a run fell short, so the fix is obvious.

name
kind
check
api_imports_correctly
deterministic
The skill's implementation in state/lib/whatsapp.ts correctly imports and uses the Meta Cloud API for WhatsApp.
send_message_functional
judge
The sendMessage() function successfully sends a WhatsApp message as intended by the input sendMessage_input.
notify_robert_functional
judge
The notifyRobert() function successfully sends a WhatsApp notification to Robert as intended by the input notifyRobert_input.
log_pending_eval_row
deterministic
A new row is appended to state/log/pending-eval.ndjson for each execution of the skill.

how it runs - the shared frame every skill uses 5/5 present

Every skill runs the same way. One part does the work, a separate part checks it, and a short loader hands the AI exactly what it needs for the job. Anything this skill doesn't use shows a one-line note saying why, on purpose, not by accident.

makes the work The worker
present
Exported functions in state/lib/whatsapp.ts. the worker
Does the actual work. Whatever it produces is what gets checked next.
checks the work The reviewer
present
None wired yet - eval is manual (Robert review). the checker
A separate checker grades the work, so the part that made it can't approve its own work.
frame
learns Self-correction
present
fixes itself learns from gaps
When a run hits a gap, the skill gets edited on the spot [FIXED] or queued for a bigger rewrite [LOGGED], so it keeps getting better.
tidies up Background fixes
present
queued for rewrite runs in the background
Bigger fixes that can't be made on the spot get queued and rewritten in the background later.
remembers Run history
present
state/log/pending-eval.ndjson pending runs
Every run is written down here, then reviewed by hand each week.
Critical rules the things this skill must not get wrong
  1. ALWAYS run voice.checkTone() over outgoing WhatsApp text BEFORE sending — em-dashes / hype / personal-story openers HARD-BLOCK
  2. Default to apply: false on send actions; require explicit apply for sendMessage() / sendMedia() — preview first
  3. Round-trip verify after send — actor (Meta API write) ≠ auditor (independent fetch via getRecentWhatsAppMessages() confirms the message exists with matching body)
  4. notifyRobert() is the privileged personal-channel send — voice gate STILL applies; Robert is not exempt
  5. Eval is manual (Robert review) — every run files to state/log/pending-eval.ndjson

what it has learned - fixes written back in over time sample

When a run hits something this skill didn't handle, the fix gets written back into the skill so it doesn't happen again. FIXED means it was corrected on the spot. LOGGED means it's queued for a bigger rewrite. Either way, the skill gets a little better and never makes the same mistake twice.

  1. Loading feedback rows…

how the work flows- who makes it, who checks it

inputs settings
actor Exported functions in state/lib/whatsapp.ts.
auditor None wired yet - eval is manual (Robert review).
1 data
eval log
`state/log/pending-eval.ndjson` (manual)

SKILL.md- the skill, written out in plain English

whatsapp

WhatsApp messaging via Meta Cloud API for all snappy-* skills.

Ported from kernel snappy-whatsapp in Phase 0.5. See state/lib/whatsapp.ts for the full API surface.

Steps

  • sendMessage() - see state/lib/whatsapp.ts
  • sendMedia() - see state/lib/whatsapp.ts
  • notifyRobert() - see state/lib/whatsapp.ts
  • getRecentWhatsAppMessages() - see state/lib/whatsapp.ts
  • markWhatsAppRead() - see state/lib/whatsapp.ts
  • reactToMessage() - see state/lib/whatsapp.ts
  • logIncomingMessage() - see state/lib/whatsapp.ts

Eval

Actor: the exported functions in state/lib/whatsapp.ts. Auditor: none wired yet - eval is manual (Robert review). File a state/log/pending-eval.ndjson row on each run.

Score convention:

OutcomeScore
Pass on first try1.0
Failed first, auto-fix applied, re-check passed0.5
Still failing or unrecoverable0.0

Gotchas

via the Phase 0.5 driver. Only these rewrites were applied: already in state/lib/)

  1. realpathSync(process.argv[1]) CLI guard wrapped in try/catch
  • See the kernel SKILL.md for the original long-form guidance if you need it

(read-only reference at the kernel path above).

Graduation

This skill is prose. Graduate by defining a deterministic auditor and flipping eval: auto.

Rubric

criteria:
  - name: api_imports_correctly
    kind: deterministic
    check: "The skill's implementation in state/lib/whatsapp.ts correctly imports and uses the Meta Cloud API for WhatsApp."
  - name: send_message_functional
    kind: judge
    check: "The sendMessage() function successfully sends a WhatsApp message as intended by the input sendMessage_input."
  - name: notify_robert_functional
    kind: judge
    check: "The notifyRobert() function successfully sends a WhatsApp notification to Robert as intended by the input notifyRobert_input."
  - name: log_pending_eval_row
    kind: deterministic
    check: "A new row is appended to state/log/pending-eval.ndjson for each execution of the skill."

AGENTS.md- what the AI loads when this skill comes up

whatsapp - loader

Per-turn rules for the whatsapp skill. Full reference: state/skills/whatsapp/SKILL.md. Do not skip these.

Critical Rules

  • ALWAYS run voice.checkTone() over outgoing WhatsApp text BEFORE sending - em-dashes / hype / personal-story openers HARD-BLOCK
  • Default to apply: false on send actions; require explicit apply for sendMessage() / sendMedia() - preview first
  • Round-trip verify after send - actor (Meta API write) ≠ auditor (independent fetch via getRecentWhatsAppMessages() confirms the message exists with matching body)
  • notifyRobert() is the privileged personal-channel send - voice gate STILL applies; Robert is not exempt
  • Eval is manual (Robert review) - every run files to state/log/pending-eval.ndjson

Commands

| ui dashboard | state/skills/whatsapp/resources/ui.openui | |library: state/lib/whatsapp.ts - sendMessage(), sendMedia(), notifyRobert(), getRecentWhatsAppMessages(), markWhatsAppRead(), reactToMessage(), logIncomingMessage() |voice gate: state/lib/voice.ts - checkTone() |eval log: state/log/pending-eval.ndjson (manual)

OpenUI Resource

  • Skill-owned OpenUI Lang resource: state/skills/whatsapp/resources/ui.openui. Read it before rendering or editing this skill's generated component surface.
  • Treat this resource as a first-class artifact of the skill, not a generic chat response. Improve it when the skill's user-facing output needs to become richer.
  • System resources compose OpenUI primitives and inherit SnappyChat tokens. Use ui_contract: branded in SKILL.md only for deliberate platform or client visuals.

Known Pitfalls

  • Treating notifyRobert() as a casual debug channel and skipping the voice gate ships AI-tells to the operator
  • Trusting the Meta API write response as proof of delivery - the round-trip fetch is the real auditor

Self-Test

An agent reading this should correctly:

  1. [ ] Voice-gate every outbound message, including notifyRobert()?
  2. [ ] Default to scope-only on send actions and require explicit apply?
  3. [ ] Round-trip verify via getRecentWhatsAppMessages() instead of trusting the write response?

Self-report

If this loader fell short, append a line:

echo "[$(date -u +%FT%TZ)] whatsapp: <what was missing>" >> state/log/loader-feedback.log

<!-- kernel-ok: historical attribution for the PID-loader pattern. -->

Self-correcting loader (PID feedback)

This is the rule that makes loaders learn. Ported from snappy-kernel, where it demonstrably worked: the snappy-dom-cartographer skill mapped Skool admin UI once, every later ship-agent inherited the full map for free.

Before you finish, do two things:

1. Fix gaps (P - proportional). MANDATORY. If this loader didn't cover your case - if you had to read another file, run an undocumented command, work around a wrong selector, or discover a quirk - you MUST attempt an Edit to this AGENTS.md before you log.

  • Default action: edit this file. Surgical. Add the missing line, table

row, or code example where it belongs. No restructuring.

  • Only LOGGED is allowed when: the fix needs >10 lines, spans multiple

files, or requires a structural rewrite. In that case the state/regen/drain.sh queue picks it up asynchronously.

  • "I didn't have time" / "it's minor" / "the next agent will figure it out"

are NOT valid reasons. Minor gaps compound - that's the whole point of this loop.

  • Concrete example of a valid P-fix: discover agent-browser screenshot

takes path positionally not as --output <path>. One line in a table. Edit it in. Do not just log it.

  • The goal: the next agent never has to leave the loader.

2. Log the result.

echo "[$(date -u +%FT%TZ)] whatsapp: <what was missing or fixed> [FIXED|LOGGED]" >> state/log/loader-feedback.log
  • FIXED = you patched this loader inline (P-fix).
  • LOGGED = too large for inline; the PostToolUse enqueue + Stop-hook drain

will rewrite the loader from scratch on next session-end.

Do not skip this. Every agent run must leave the system better than it found it. The loader is the setpoint; you are the sensor; the gap is the error signal; closing the gap is the correction.

api.ts- the code it can call

#!/usr/bin/env npx tsx
/**
 * snappy-whatsapp/api.ts -- WhatsApp messaging via Meta Cloud API for all snappy-* skills.
 *
 * Direct WhatsApp Cloud API calls -- no Xano middleware.
 * Phone numbers MUST be E.164 format (+14155551212).
 *
 * Required env vars (via snappy-settings/.env.cache):
 *   WHATSAPP_TOKEN    -- Meta Business API access token
 *   WHATSAPP_PHONE_ID -- WhatsApp Business phone number ID
 *   ROBERT_PHONE      -- Robert's phone for notifyRobert (E.164)
 *
 * Usage:
 *   npx tsx api.ts send "+14155551212" "Hey, quick update..."
 *   npx tsx api.ts media "+14155551212" "https://example.com/img.png" "Caption"
 *   npx tsx api.ts notify "Build completed"
 *
 * Or import as module:
 *   import { sendMessage, sendMedia, notifyRobert, getRecentWhatsAppMessages } from "./whatsapp.ts";
 */

import { env } from "./env.ts";
import { realpathSync } from "fs";

const GRAPH_API = "https://graph.facebook.com/v21.0";

const CONFIG_ERROR =
  "snappy-whatsapp not configured — add WHATSAPP_TOKEN, WHATSAPP_PHONE_ID, ROBERT_PHONE to .env.cache";

function requireConfig(): { token: string; phoneId: string; robertPhone: string } {
  const t = env("WHATSAPP_TOKEN", false);
  const p = env("WHATSAPP_PHONE_ID", false);
  const r = env("ROBERT_PHONE", false);
  if (!t || !p || !r) throw new Error(CONFIG_ERROR);
  return { token: t, phoneId: p, robertPhone: r };
}

async function whatsappFetch(body: Record<string, unknown>) {
  const { token, phoneId } = requireConfig();
  const res = await fetch(`${GRAPH_API}/${phoneId}/messages`, {
    method: "POST",
    headers: {
      Authorization: `Bearer ${token}`,
      "Content-Type": "application/json",
    },
    body: JSON.stringify({ messaging_product: "whatsapp", ...body }),
  });
  const data = await res.json();
  if (!res.ok) {
    throw new Error(`WhatsApp Cloud API failed (${res.status}): ${JSON.stringify(data)}`);
  }
  return data;
}

// --- Public API ---

export async function sendMessage(to: string, message: string) {
  return whatsappFetch({
    to,
    type: "text",
    text: { body: message },
  });
}

export async function sendMedia(to: string, mediaUrl: string, caption?: string) {
  return whatsappFetch({
    to,
    type: "image",
    image: { link: mediaUrl, ...(caption ? { caption } : {}) },
  });
}

export async function notifyRobert(text: string) {
  const { robertPhone } = requireConfig();
  return sendMessage(robertPhone, text);
}

// --- Read operations (webhook-based) ---

/**
 * Fetch recent messages from the WhatsApp webhook log.
 * WhatsApp Cloud API doesn't support polling — messages arrive via webhook.
 * This reads from the local webhook log file if it exists.
 */
export async function getRecentWhatsAppMessages(limit = 20): Promise<Array<{
  from: string;
  text: string;
  timestamp: string;
  message_id: string;
  type: string;
}>> {
  const { existsSync, readFileSync } = await import("fs");
  const { join } = await import("path");
  const logPath = join(process.env.HOME!, ".claude/skills/snappy-whatsapp/webhook-log.jsonl");
  if (!existsSync(logPath)) return [];
  const lines = readFileSync(logPath, "utf-8").trim().split("\n").filter(Boolean);
  return lines.slice(-limit).map(line => {
    const entry = JSON.parse(line);
    return {
      from: entry.from || "unknown",
      text: entry.text || entry.caption || "[non-text]",
      timestamp: entry.timestamp || "",
      message_id: entry.message_id || "",
      type: entry.type || "text",
    };
  });
}

/**
 * Mark a message as read (sends read receipt).
 */
export async function markWhatsAppRead(messageId: string) {
  const { token, phoneId } = requireConfig();
  const res = await fetch(`${GRAPH_API}/${phoneId}/messages`, {
    method: "POST",
    headers: {
      Authorization: `Bearer ${token}`,
      "Content-Type": "application/json",
    },
    body: JSON.stringify({
      messaging_product: "whatsapp",
      status: "read",
      message_id: messageId,
    }),
  });
  return res.json();
}

/**
 * React to a message with an emoji.
 */
export async function reactToMessage(messageId: string, emoji: string) {
  return whatsappFetch({
    to: "", // not needed for reactions but required by schema
    type: "reaction",
    reaction: { message_id: messageId, emoji },
  });
}

/**
 * Log an incoming webhook message to the local log file.
 * Call this from the webhook handler to build the read log.
 */
export async function logIncomingMessage(payload: Record<string, unknown>) {
  const { appendFileSync, mkdirSync } = await import("fs");
  const { join, dirname } = await import("path");
  const logPath = join(process.env.HOME!, ".claude/skills/snappy-whatsapp/webhook-log.jsonl");
  mkdirSync(dirname(logPath), { recursive: true });
  appendFileSync(logPath, JSON.stringify({ ...payload, logged_at: new Date().toISOString() }) + "\n");
}

// --- CLI ---

if ((() => { try { return import.meta.url === `file://${realpathSync(process.argv[1])}`; } catch { return false; } })()) {
  (async () => {
    const [, , cmd, ...args] = process.argv;

    switch (cmd) {
      case "send": {
        const [to, ...msgParts] = args;
        if (!to || !msgParts.length) { console.error("Usage: api.ts send <+E.164> <message>"); process.exit(1); }
        const data = await sendMessage(to, msgParts.join(" "));
        console.log(JSON.stringify(data, null, 2));
        break;
      }
      case "media": {
        const [to, mediaUrl, ...captionParts] = args;
        if (!to || !mediaUrl) { console.error("Usage: api.ts media <+E.164> <url> [caption]"); process.exit(1); }
        const data = await sendMedia(to, mediaUrl, captionParts.join(" ") || undefined);
        console.log(JSON.stringify(data, null, 2));
        break;
      }
      case "notify": {
        const text = args.join(" ");
        if (!text) { console.error("Usage: api.ts notify <text>"); process.exit(1); }
        const data = await notifyRobert(text);
        console.log(JSON.stringify(data, null, 2));
        break;
      }
      case "read": {
        const limit = args[0] ? parseInt(args[0], 10) : 20;
        const msgs = await getRecentWhatsAppMessages(limit);
        if (msgs.length === 0) {
          console.log("No messages in webhook log. Set up webhook handler to log incoming messages.");
        } else {
          for (const m of msgs) {
            console.log(`${m.timestamp}\t${m.from}\t${m.text.slice(0, 300)}`);
          }
        }
        break;
      }
      case "mark-read": {
        const [msgId] = args;
        if (!msgId) { console.error("Usage: api.ts mark-read <message_id>"); process.exit(1); }
        await markWhatsAppRead(msgId);
        console.log("marked read");
        break;
      }
      default:
        console.log("Usage: npx tsx api.ts [send|media|notify|read|mark-read] ...");
    }
  })();
}

scripts- helper scripts it can run

prose-only skill - 1 inline code block live in SKILL.md above (no state/bin/ sidecar yet).

how we check it- the checks, plus the last 10 runs

rubric shape schema-shape check (no inline rubric)
recent mean 1.00 · 10 runs actor/auditor: unverifiable
deps settings
timestamp verb score primary_issue artifact
2026-04-25 04:11Z - 1.00 - -
2026-04-21 15:58Z - 1.00 - -
2026-04-21 15:57Z - 1.00 - -
2026-04-21 03:53Z - 1.00 - -
2026-04-25 04:11Z - 1.00 - -
2026-04-21 15:58Z - 1.00 - -
2026-04-21 15:57Z - 1.00 - -
2026-04-21 03:53Z - 1.00 - -
2026-04-25 04:11Z - 1.00 - -
2026-04-21 15:58Z - 1.00 - -