agentleFS
korchard333korchard333/claude-power-platform-community/.claude/skills/ai-builder/SKILL.md

ai-builder

Skill46 starsChanged 8 months agoMITView on GitHub

AI Builder for Power Platform. Use when: document processing, prediction models, GPT prompt actions in flows, prebuilt models (text recognition, sentiment, entity extraction), custom models, AI Builder in Canvas Apps and Power Automate, licensing.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Not testing models with documents that differ from training samples

    line 56
  2. Distinctive:

    Not monitoring AI Builder credit consumption (can exhaust capacity unexpectedly)

    line 59
  3. Distinctive:

    No error handling around AI Builder actions in flows (models can fail on unexpected input)

    line 62

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Skill: AI Builder

When to Use

Trigger when building, configuring, or integrating AI Builder models -- document processing, prediction, GPT prompts, prebuilt models, or custom ML models within Power Platform.


Model Types

Model TypeDescriptionTraining Required?
Document processingExtract fields from invoices, receipts, custom documentsYes -- train with sample documents
Text recognition (OCR)Extract text from images and PDFsNo (prebuilt)
Sentiment analysisDetect positive/negative/neutral sentiment in textNo (prebuilt)
Entity extractionExtract entities (names, dates, addresses) from textNo (prebuilt)
Text classificationCategorize text into custom categoriesYes -- train with labeled examples
Object detectionDetect and count objects in imagesYes -- train with labeled images
PredictionPredict outcomes based on historical dataYes -- requires Dataverse table with history
GPT promptsCustom prompts using Azure OpenAI GPT modelsNo -- prompt engineering only
Category classificationClassify items into predefined categoriesYes

⚠️ REQUIRED: Load Sub-Files Before Implementation

SKILL.md is a summary only — it is NOT sufficient for implementation.

The detailed content (complete payloads, XML templates, working examples, edge-case handling) lives in sub-files in the same directory as this SKILL.md. Before writing any code, you MUST use read_file on the sub-files relevant to your task:

  • Document Processing -- Document processing models, training, publishing, Power Automate integration, batch processing, confidence thresholds, human review loop
  • GPT Prompts -- GPT prompt actions, prompt engineering for Power Platform, Power Automate and Canvas App integration, token management
  • Predictions -- Prediction models, category classification, object detection, training data requirements, scheduling retraining, performance monitoring

Licensing

CapabilityLicense Requirement
Prebuilt models (OCR, sentiment)Included with Power Apps/Automate Premium
Custom models (document processing, prediction)AI Builder add-on capacity (credits)
GPT promptsAI Builder add-on capacity (credits)
ConsumptionCredit-based -- different models consume different credits per call

Tip: Start with prebuilt models (no additional cost with Premium license) before investing in custom model training.


Anti-Patterns

  • Training document processing models with fewer than 5 sample documents (poor accuracy)
  • Not testing models with documents that differ from training samples
  • Using GPT prompts for deterministic logic (use expressions or business rules instead)
  • Hardcoding model IDs in flows -- use environment variables for environment promotion
  • Not monitoring AI Builder credit consumption (can exhaust capacity unexpectedly)
  • Processing sensitive data through GPT prompts without reviewing data residency policies
  • Using custom models when a prebuilt model already handles the document type
  • No error handling around AI Builder actions in flows (models can fail on unexpected input)

Related Skills

  • power-automate -- AI Builder actions are most commonly used in cloud flows
  • canvas-apps -- AI Builder components for in-app document processing and predictions
  • dataverse -- Prediction models train on Dataverse table data; results stored in Dataverse
  • alm -- AI Builder models are solution-aware and promote with managed solutions
  • copilot-studio -- GPT prompts can be used as plugin actions in Copilot agents
  • azure-openai -- When AI Builder GPT limits are insufficient, escalate to direct Azure OpenAI

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.