All loops
FinanceHard 12 min· claude-sonnet-4-5

Spend categorizer

Spend categorizer with auditable Claude reasoning.

Read the full Finance build guide →
NOT DEPLOYEDNOT DEPLOYED
0160ms
Trigger
cron(0 7 * * *) fired · every day · 07:00
02940ms
Agent
claude-sonnet-4-5 · in 1240 tok · out 397 tok
03210ms
Tools
postgres-mcp/stripe:customers.retrieve → 200 OK · 180ms
0450ms
Verify
schema check · zod v3 passed
0580ms
Output
slack #finance · anomaly flagged
0640ms
Notify
audit log written · runbook link attached
SUCCESS
0%
0 runs
P50
0ms
median
P95
0ms
tail
AVG COST
per run
LAST OK
never
LAST FAIL
never
none
Latency · last 30 runs0 samples
no runs yet
Latest output · what your users see
Invoice · AnthropicInfra
$1,060.30 USD
⚠ anomaly · +312% vs 30-day avg
// press Test to run once · Watch live to keep streaming · Deploy to make it real
The problem

Finance ops are spreadsheets glued together with hope.

The outcome

A scheduled agent does the join, the variance, and the narrative.

Ingredients & skills

Secrets
  • ANTHROPIC_API_KEY
  • STRIPE_KEY
Providers
  • Anthropic
  • Stripe
MCP servers
  • postgres-mcp
  • fetch-mcp
#finance#ops#claude

How it works

Spend categorizer with auditable Claude reasoning.

Step 1

1 — Pull

Read from billing + GL via read-only keys.

Step 2

2 — Join + narrate

Claude joins, computes, and writes the narrative with citations.

Step 3

3 — Distribute

PDF + Slack + (optional) board portal.

One-line deploy

The button above runs the same command with your saved config. This is the raw CLI form.

bash
npx claudeloops deploy spend-categorizer

Related loops