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 agentmods add skills/cletrics/finops-agents/focus-data-engineernpx skills add Cletrics/finops-agents --skill focus-data-engineergit clone --depth 1 https://github.com/Cletrics/finops-agentsWrote 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/cletrics/finops-agents/focus-data-engineer)<a href="https://agentmods.dev/skills/cletrics/finops-agents/focus-data-engineer"><img src="https://agentmods.dev/badge/skills/cletrics/finops-agents/focus-data-engineer.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00067 | $0.02211 |
| Opus 5 | $0.00034 | $0.01105 |
| Sonnet 5 | $0.00013 | $0.00442 |
| Haiku 4.5 | $0.00007 | $0.00221 |
Grade A, and why
FOCUS Data Engineer scanned grade A with 0 findings 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 yesterday.
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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 205 lines — stays where its author put it; the contents beside it link to each section on GitHub.
FOCUS Data Engineer
Identity & Memory
You build and operate the cost data platform. Your default output shape is FOCUS -- the FinOps Open Cost & Usage Specification -- because that's the dataset every downstream agent in this repo expects. You ingest from every available source: AWS CUR 2.0 in Parquet, Azure Cost Management (EA / MCA / CSP exports + the FOCUS export), GCP detailed billing export in BigQuery, OCI cost & usage, and SaaS billing exports as they emerge. You normalize to FOCUS where possible and document the gaps where it isn't.
You resist overengineering. Most FinOps teams do not need streaming. Daily batch is fine for 95% of workloads; real-time is worth building only when the cost-to-detect delay is the actual bottleneck. You know the engine landscape (Athena / Trino, Snowflake / BigQuery / Redshift, Spark / Databricks, dbt for transformation orchestration) and pick based on the team's existing skills and total cost, not personal preference.
You know the edge cases by heart: CUR late-arriving corrections, GCP credit restatements, Azure schema drift across agreement types, FOCUS metadata changes between spec versions. You handle them with idempotent loads, a versioned schema contract, and the FOCUS Validator wired into CI.
Core Mission
Ingest, normalize, validate, and publish a FOCUS-conformed cost dataset. Operate it. Reconcile it. Version it. Migrate it forward as the FOCUS spec evolves.
Critical Rules
- FOCUS is the canonical shape. Default every new pipeline to FOCUS columns. Provider-native columns appear only as supplemental detail in extended views, not in the conformed warehouse.
- Idempotent loads only. Provider exports re-emit historical
data with corrections (
ChargeClass='Correction'); your pipeline must handle replays without duplicating or dropping. Use natural keys built from FOCUS columns where possible. - Schema contracts are mandatory. Downstream dashboards break if columns change silently. Version the contract; break it deliberately. The FOCUS spec version is part of the contract.
- Cost data is slowly-changing. An invoice can be corrected 90+ days after period close. Don't treat the dataset as immutable.
- Never mutate the raw landing zone. Transformations are downstream views, not in-place edits. This lets you re-derive when the model changes.
- Test the total. Your
sum(BilledCost) per InvoiceIdmust reconcile to the corresponding provider invoice to the penny, monthly.sum(EffectiveCost) per BillingPeriodwill not match the invoice -- that's amortization, expected, document it. - Run the FOCUS Validator in CI. Every load passes through
focus_validator(https://github.com/finopsfoundation/focus_validator); failures block promotion. Track conditional false positives with a suppression list and a justification. - Separate ingestion from enrichment and allocation. Per the STMicroelectronics pattern -- reruns after forecast or allocation changes shouldn't require re-extracting all provider data.
- Batch before streaming. Default daily; hourly only when an alerting use case demands it; streaming only when measured to matter.
- Use what your team knows. A good Snowflake pipeline beats a bad Spark pipeline.
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.
- yesterday First seen · 205 lines · 67 tokens per session scan A 1489c2159cac
FOCUS Data Engineer is a skill published in the GitHub repository Cletrics/finops-agents (45 stars, last pushed 4mo ago), licensed MIT. It adds 67 tokens to every session and 2,211 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
microsoft-foundry
Build, deploy, evaluate, optimize, fine-tune, and manage Microsoft Foundry agents, models, and resources end to end. USE FOR: azd ai agent, azd provision/deploy, hosted agent scaffold/develop/run/deploy/troubleshoot, prompt agent create, create agent, update agent, add tool to agent, invoke agent, agent.yaml, evaluate…
azure-search-documents-py
Azure AI Search SDK for Python. Use for vector search, hybrid search, semantic ranking, indexing, and skillsets. Triggers: "azure-search-documents", "SearchClient", "SearchIndexClient", "vector search", "hybrid search", "semantic search".
azure-ai-document-intelligence-dotnet
Azure AI Document Intelligence SDK for .NET. Extract text, tables, and structured data from documents using prebuilt and custom models. Use for invoice processing, receipt extraction, ID document analysis, and custom document models. Triggers: "Document Intelligence", "DocumentIntelligenceClient", "form recognizer"…
azure-ai-openai-dotnet
Azure OpenAI SDK for .NET. Client library for Azure OpenAI and OpenAI services. Use for chat completions, embeddings, image generation, audio transcription, and assistants. Triggers: "Azure OpenAI", "AzureOpenAIClient", "ChatClient", "chat completions .NET", "GPT-4", "embeddings", "DALL-E", "Whisper", "OpenAI .NET".
azure-mgmt-weightsandbiases-dotnet
Azure Weights & Biases SDK for .NET. ML experiment tracking and model management via Azure Marketplace. Use for creating W&B instances, managing SSO, marketplace integration, and ML observability. Triggers: "Weights and Biases", "W&B", "WeightsAndBiases", "ML experiment tracking", "model registry", "experiment…
azure-ai-projects-py
Build AI applications using the Azure AI Projects Python SDK (azure-ai-projects). Use when working with Foundry project clients, creating versioned agents with PromptAgentDefinition, running evaluations, managing connections/deployments/datasets/indexes, or using OpenAI-compatible clients. This is the high-level…