Exported functions in state/lib/inbound.ts. .md file to compare - side-by-side diff against inbound
inbound
description: "Triggers on prompt mention of 'inbound'."
What it does for you
Spots and sorts incoming messages so the right ones reach you.
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.
For developers how this skill is built, graded, and how it runs
at a glance- the short version
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.
state/skills/inbound/SKILL.md
present
state/lib/inbound.ts
present
state/bin/inbound/
not present
state/skills/inbound/AGENTS.md
present
how it's graded - what counts as a good run 4 criteria · 3 deterministic · 1 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.
how it runs - the shared frame every skill uses 4/5 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.
This skill doesn't fix its own gaps yet.
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.
- Loading feedback rows…
how the work flows- who makes it, who checks it
import functions from `state/lib/inbound.ts` (`checkSlackJoins`, `checkEmail`, `classifyInbound`)
SKILL.md- the skill, written out in plain English
inbound
Inbound detection and classification.
Ported from kernel snappy-inbound in Phase 0.5. See state/lib/inbound.ts for the full API surface.
Steps
checkSlackJoins()- seestate/lib/inbound.tscheckEmail()- seestate/lib/inbound.tsclassifyInbound()- seestate/lib/inbound.ts
Eval
Actor: the exported functions in state/lib/inbound.ts. Auditor: none wired yet - eval is manual (Robert review). File a state/log/pending-eval.ndjson row on each run.
Score convention:
| Outcome | Score |
|---|---|
| Pass on first try | 1.0 |
| Failed first, auto-fix applied, re-check passed | 0.5 |
| Still failing or unrecoverable | 0.0 |
Gotchas
via the Phase 0.5 driver. Only these rewrites were applied: already in state/lib/)
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: calls_all_required_functions
kind: deterministic
check: "The execution log includes calls to checkSlackJoins(), checkEmail(), and classifyInbound()."
- name: inputs_are_strings
kind: deterministic
check: "Inputs checkSlackJoins_input, checkEmail_input, and classifyInbound_input are all provided as strings."
- name: creates_log_row
kind: deterministic
check: "A new row is appended to state/log/pending-eval.ndjson after skill execution."
- name: inbound_classification_valid
kind: judge
check: "The output of classifyInbound() aligns with expected inbound classification based on provided inputs."AGENTS.md- what the AI loads when this skill comes up
inbound - loader
Per-turn rules for the inbound skill. Full reference: state/skills/inbound/SKILL.md. Do not skip these.
Critical Rules
_(no failures recorded yet - this skill is a Phase 0.5 mechanical port from snappy-inbound with no hard-won rules. Read state/skills/inbound/SKILL.md and state/lib/inbound.ts before invoking. Eval is manual until an auditor is wired.)_
Commands
| ui model | live composition via compose_inline, persisted as artifact lang_body, reopened with OpenArtifact | |invoke: import functions from state/lib/inbound.ts (checkSlackJoins, checkEmail, classifyInbound) |eval log: state/log/pending-eval.ndjson (manual review until shape gate added)
Known Pitfalls
- Skill is
eval: shapein frontmatter but auditor is "none wired yet - manual" - every run musteval.pending()row, do not claim auto-pass
Self-Test
An agent reading this should correctly:
- [ ] Log every run to
state/log/pending-eval.ndjsonrather thanevals.ndjson? - [ ] Resist drafting a deterministic auditor without first running the function on real data twice?
- [ ] Find the API surface in
state/lib/inbound.ts?
Found a gap? Edit this file. <!-- footer-injection-point -->
api.ts- the code it can call
#!/usr/bin/env npx tsx
/**
* snappy-inbound/api.ts -- Inbound detection and classification.
*
* Usage:
* npx tsx api.ts slack-joins # check #all-snappy for join events
* npx tsx api.ts classify "I want to learn about AI agents for my business"
*
* Or import as module:
* import { checkSlackJoins, classifyInbound } from "./inbound.ts";
*/
import { readMessages } from "./slack.ts";
import { env } from "./env.ts";
import { realpathSync } from "fs";
const ALL_SNAPPY_CHANNEL = "C09DD2D0S07";
/** Check #all-snappy for recent join events. Returns messages with join subtype. */
export async function checkSlackJoins(limit = 50) {
const data = await readMessages(ALL_SNAPPY_CHANNEL, limit);
if (!data.messages) return [];
return data.messages.filter(
(m: any) => m.subtype === "channel_join" || m.subtype === "group_join"
);
}
/**
* Check email for new subscribers.
* Placeholder -- ActiveCampaign is NOT in use. Email goes through Xano/Gmail.
* When implementing, use Xano contacts API or Loops.so for subscriber list polling.
*/
export async function checkEmail(): Promise<{ status: string; info: string }> {
return {
status: "not_implemented",
info: "Subscriber polling not yet implemented. Use Xano contacts API (GET api:PB9UH7b9/contacts) or Loops.so. ActiveCampaign is NOT in use.",
};
}
type InboundType = "lead" | "support" | "spam" | "partnership" | "unknown";
/** Classify inbound text into a type. Simple keyword-based -- upgrade to LLM call if needed. */
export function classifyInbound(text: string): { type: InboundType; confidence: number } {
const lower = text.toLowerCase();
const spamSignals = ["unsubscribe", "casino", "crypto airdrop", "free money", "click here now"];
if (spamSignals.some((s) => lower.includes(s))) {
return { type: "spam", confidence: 0.9 };
}
const partnerSignals = ["partnership", "collaborate", "joint venture", "co-market", "affiliate"];
if (partnerSignals.some((s) => lower.includes(s))) {
return { type: "partnership", confidence: 0.7 };
}
const supportSignals = ["help", "issue", "broken", "not working", "bug", "error", "can't access"];
if (supportSignals.some((s) => lower.includes(s))) {
return { type: "support", confidence: 0.7 };
}
const leadSignals = [
"interested", "pricing", "how much", "learn more", "demo",
"business", "agency", "consulting", "ai agent", "automation",
];
if (leadSignals.some((s) => lower.includes(s))) {
return { type: "lead", confidence: 0.7 };
}
return { type: "unknown", confidence: 0.3 };
}
// --- CLI ---
if ((() => { try { return import.meta.url === `file://${realpathSync(process.argv[1])}`; } catch { return false; } })()) {
(async () => {
const [, , cmd, ...args] = process.argv;
switch (cmd) {
case "slack-joins": {
const joins = await checkSlackJoins();
if (joins.length === 0) {
console.log("No recent join events.");
} else {
for (const j of joins) {
const ts = new Date(Number(j.ts) * 1000).toISOString().slice(0, 16);
console.log(`${ts}\t${j.user}\tjoined`);
}
}
break;
}
case "classify": {
const text = args.join(" ");
if (!text) { console.error("Usage: api.ts classify <text>"); process.exit(1); }
const result = classifyInbound(text);
console.log(JSON.stringify(result, null, 2));
break;
}
default:
console.log("Usage: npx tsx api.ts [slack-joins|classify] ...");
}
})();
}
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
no recent runs logged - the eval contract is declared but nothing has been graded yet