economics-provider-collection

Collect and reconcile cloud provider balances, usage, and invoices for monthly close.

5.0k|1.0k|Updated Apr 15, 2021
One-click install
npx skills add https://github.com/pollinations/pollinations --skill economics-provider-collection
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: economics-provider-collection
Source: https://github.com/pollinations/pollinations/tree/main/.claude/skills/economics-provider-collection
Command: npx skills add https://github.com/pollinations/pollinations --skill economics-provider-collection

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Finance and operations teams need accurate monthly cost data from dozens of AI and cloud vendors, but each provider exposes billing through different APIs, CLIs, dashboards, and invoice formats. This Skill standardizes the collection and reconciliation of vendor balances, usage, and invoice evidence into consistent ledgers.

Core Features & Use Cases

  • Registry-driven vendor collection: Reads a canonical provider registry to determine which vendors are active, their accounts, connectors, and access URLs before collecting data.
  • Multi-source evidence gathering: Uses provider APIs, CLIs, MCP connectors, or authenticated dashboards in priority order, with per-vendor guides for AWS, Azure, Google Cloud, Anthropic, OpenAI, Cloudflare, and many more.
  • Ledger reconciliation: Writes verified provider usage, balances, and bank movements into economics ledgers with deterministic entry IDs, backup verification, and approval-gated writes.
  • Use Case: During a monthly close, refresh all active vendors, collect last month's usage at model or SKU grain, archive invoices to Google Drive, and reconcile provider costs against internal Pollen consumption.

Quick Start

Ask the agent to refresh all active vendors for last month and reconcile provider usage against the economics ledgers.

Frequently Asked Questions about economics-provider-collection

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I collect monthly usage and invoices from multiple cloud providers?

Read the provider registry first to identify active vendors, then follow each vendor's connector guide to query its API, CLI, or dashboard in priority order. Save raw evidence to the data inbox and archive invoices in Google Drive before writing ledger entries.

What is the correct order for collecting provider billing data?

Use the first source that gives authoritative, granular data: supported provider API, then supported CLI, then MCP or authenticated connector, then the provider dashboard. Do not build permanent integrations for a bounded monthly task.

Can I use provider balance snapshots as monthly usage evidence?

No. A balance snapshot is current state, not historical burn, and a top-up is funding rather than usage. Monthly usage must come from billing APIs, invoices, or dashboard usage views bounded to the closed calendar month.

How are corrections to existing ledger entries handled?

Reuse the original deterministic entry_id when correcting a fact, include the current base version, use a later recorded_at timestamp, and preserve immutable identity fields. Never create a second identity for the same fact.

What happens when a provider API lacks model-level granularity?

Preserve the exact provider total and explicitly state the missing granularity. Never allocate a total using internal model names or undocumented ratios; keep the gap visible in reconciliation notes.

Does this skill create or rotate provider credentials during collection?

No. Collection uses only existing authorized credentials. Creating, rotating, or deploying credentials requires the repository's separate secret-mutation plan and explicit scoped approval.