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 skills add wardawgmalvicious/agent-config --skill fabric-data-agentgit clone --depth 1 https://github.com/wardawgmalvicious/agent-configWrote 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/wardawgmalvicious/agent-config/fabric-data-agent)<a href="https://agentmods.dev/skills/wardawgmalvicious/agent-config/fabric-data-agent"><img src="https://agentmods.dev/badge/skills/wardawgmalvicious/agent-config/fabric-data-agent/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/wardawgmalvicious/agent-config/fabric-data-agent"><img src="https://agentmods.dev/badge/skills/wardawgmalvicious/agent-config/fabric-data-agent.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00230 | $0.02726 |
| Opus 5 | $0.00115 | $0.01363 |
| Sonnet 5 | $0.00046 | $0.00545 |
| Haiku 4.5 | $0.00023 | $0.00273 |
Grade A, and why
fabric-data-agent 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 4d ago.
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 — 135 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Configuring Fabric Data Agents
A practical, reusable guide for configuring a Fabric Data Agent so it returns accurate, contextually relevant answers. Use this template across projects — replace the example domain (retail / sales / logistics) with your own without changing the structure.
What a Data Agent is
A Fabric Data Agent is a conversational Q&A interface. It accepts natural-language questions, routes them to the right data source, generates a query (SQL / DAX / KQL / Microsoft Graph), validates it, executes it read-only, and returns a human-readable answer.
Supported data sources: Lakehouse, Warehouse, KQL Database (Eventhouse), Power BI Semantic Model, Fabric SQL Database, Mirrored Database, Ontology, Microsoft Graph. A single agent supports up to 5 data sources in any combination. Read-only by design — it never generates create/update/delete queries.
Ontology as a source is the one entry there that is a whole item type of its own — see fabric-ontology for modelling it, binding it to data, and its semantic enrichment, which is what makes an ontology-grounded agent answer well. Three data-agent-side behaviours belong here rather than there: an ontology source is still preview, the agent's first few queries after creation can fail while it initializes (wait and retry), and aggregation is a known gap — add the instruction Support group by in GQL to the agent's instructions.
GA since March 2026 for the core surface: create / configure / publish / share, built-in diagnostics, and lifecycle management via Git integration + deployment pipelines. Everything else named in this skill as preview — Creator Agent, MCP endpoint, M365 Copilot, Python SDK, Copilot Studio, Foundry, SPN auth — should be gated behind an explicit decision. Two integration paths retired 2026-08-26: the Azure OpenAI Assistants API (migrate to the MCP endpoint; within the SDK, to the Fabric OpenAI Responses client) and Copilot in Power BI. Full status breakdown and migration detail: references/status-and-retirements.md.
What ships with it
8 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- assets/example-retail-agent.md 3.6 KB
- references/authentication.md 2.5 KB
- references/authoring-workflow.md 4.7 KB
- references/configuration-layers.md 7.5 KB
- references/consumption-surfaces.md 3.6 KB
- references/creator-agent.md 2.1 KB
- references/REFERENCE.md 4.2 KB
- references/status-and-retirements.md 2.4 KB
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.
- 4d ago Changed · -205 lines scan B → A 58c7edaf4206
- 8d ago First seen · 340 lines · 230 tokens per session scan B 2f25439615af
fabric-data-agent is a skill published in the GitHub repository wardawgmalvicious/agent-config (1 stars, last pushed today), licensed MIT. It adds 230 tokens to every session and 2,726 once invoked, about $0.0011 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 2026-08-31.
Other skills, from other repositories
llm-app-patterns
Production-ready patterns for building LLM applications. Covers RAG pipelines, agent architectures, prompt IDEs, and LLMOps monitoring. Use when designing AI applications, implementing RAG, building agents, or setting up LLM observability.
blip-2-vision-language
Vision-language pre-training framework bridging frozen image encoders and LLMs. Use when you need image captioning, visual question answering, image-text retrieval, or multimodal chat with state-of-the-art zero-shot performance.
grpo-rl-training
Expert guidance for GRPO/RL fine-tuning with TRL for reasoning and task-specific model training.
langchain
Framework for building LLM-powered applications with agents, chains, and RAG. Supports multiple providers (OpenAI, Anthropic, Google), 500+ integrations, ReAct agents, tool calling, memory management, and vector store retrieval. Use for building chatbots, question-answering systems, autonomous agents, or RAG…
knowledge-distillation
Compress large language models using knowledge distillation from teacher to student models. Use when deploying smaller models with retained performance, transferring GPT-4 capabilities to open-source models, or reducing inference costs. Covers temperature scaling, soft targets, reverse KLD, logit distillation, and…
speculative-decoding
Accelerate LLM inference using speculative decoding, Medusa multiple heads, and lookahead decoding techniques. Use when optimizing inference speed (1.5-3.6× speedup), reducing latency for real-time applications, or deploying models with limited compute. Covers draft models, tree-based attention, Jacobi iteration…