Skip to content
Skills Library
AI & agents · Building & coding

salesforce-einstein

/salesforce-einstein

Use for Salesforce Einstein work — Discovery, Predictions, Next Best Action, Einstein Bots (legacy → migrating to Agentforce), prompt builder, model management.

Repository
sift-coder ↗
Licence
MIT
Last repository change
03.06.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 03.06.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Use for Salesforce Einstein work — Discovery, Predictions, Next Best Action, Einstein Bots (legacy → migrating to Agentforce), prompt builder, model management.

Original author description

Use for Salesforce Einstein work — Discovery, Predictions, Next Best Action, Einstein Bots (legacy → migrating to Agentforce), prompt builder, model management.

How it works

  1. Surfaces
  2. Method (typical workflows)
  3. Output shape
  4. Rules
  5. Anti-patterns
  6. When NOT to use
  7. Subagent dispatch
  8. Key references
  9. Value over native CC
Original SKILL.md headings · 97 lines, 487 words
SKILL.md preview
---
name: salesforce-einstein
description: Use for Salesforce Einstein work — Discovery, Predictions, Next Best Action, Einstein Bots (legacy → migrating to Agentforce), prompt builder, model management.
paths: '**/*.cls,**/*.trigger,**/*.apex,**/lwc/**,**/aura/**,**/objects/**,**/*.object-meta.xml,sfdx-project.json,**/flows/**,**/flexipages/**,**/permissionsets/**,**/profiles/**'
---

# salesforce-einstein

Salesforce Einstein platform. Distinct from Agentforce (which subsumes some Einstein features). Use for: predictions, classifications, recommendations, prompt-builder workflows.

## Surfaces

- **Einstein Discovery** — AutoML on CRM data; predictions + model explainability
- **Einstein Prediction Builder** — clicks-not-code prediction model
- **Einstein Next Best Action** — recommendation strategy + flows
- **Prompt Builder** — prompt templates for AI features
- **Einstein Bots** — legacy chatbots; new builds → Agentforce
- **Model deployment** — Connect API for custom models, Trust Layer integration

## Method (typical workflows)

### Discovery model

1. Identify the **target field** (what to predict).
2. Identify the **comparison data** (positive vs negative outcomes).
3. Configure **dataset** — fields to include, exclusions for bias.
4. Train. Inspect **model fit** (R², accuracy, lift).
5. Deploy — embed in Lightning page or Flow.
6. **Monitor drift** — schedule re-train or alert on accuracy drop.

### Prompt template

1. **Type** — Field Generation / Email Generation / Sales Email / Flex / Record Summary.
2. **Resources** — fields, related lists, files, Apex methods, Flows.
3. **Instructions** — natural language; avoid contradicting the resource shape.
4. **Test** — Prompt Builder preview; eval against sample records.
5. **Surface** — embed in record page action, Flow, or via Apex.

## Output shape

```
Workflow:    <Discovery | Prediction Builder | NBA | Prompt | Bot>

Inputs:
  Object:     <SObject>
  Dataset:    <fields, filters>

Configuration:
  <key params>

Quality gates:
  Model fit:        <metric, threshold>
  Bias check:       <how>
  Trust Layer:      <PII handling>

Deploy:
  Surface:    <Lightning page | Flow | Apex | Bot>
  Test plan:  <eval cases>

Monitoring:
  Metric:     <accuracy, drift, latency>
  Alert:      <threshold>
```

## Rules

- **Bias review is part of quality, not optional.** Exclude biased fields (race, age, gender unless compliance-required).
- **Trust Layer applies to LLM-backed features** (prompts, generative responses).
- **Capacity** — Einstein has compute limits per edition; check before adding heavy models.
- **Migration path** — new chatbot work goes to Agentforce, not Einstein Bots.
- **Eval cases captured to memory** for regression testing.

## Anti-patterns

- Discovery model using PII as input without compliance signoff
- Prompt template that ignores resources, hallucinates from imagined records
- Deploying without monitoring (drift bites silently)
- Building new Einstein Bot when org has Agentforce licence
- Custom-model integration without Trust Layer

## When NOT to use

- Non-Einstein-licenced org — features unavailable
- Pure rule-based logic — use Flow / Apex
- Agentforce work — `/siftcoder:salesforce-agentforce`

## Subagent dispatch

- `salesforce-architect` for capacity + bias review
- `Plan` for multi-step workflows (data prep + train + deploy)
- `general-purpose` for the metadata generation

## Key references

- Einstein Discovery: trailhead.salesforce.com (Einstein Discovery Basics)
- Prompt Builder: help.salesforce.com → Prompt Builder
- Einstein Trust Layer: help.salesforce.com (also referenced from agentforce skill)

## Value over native CC

CC will discuss ML if asked. CC won't naturally know Einstein-specific shapes (Discovery model fit metrics, Prompt Builder template types, NBA strategy components). Platform-specific knowledge IS the value.

Versions

The database has no other skills named salesforce-einstein.

Database snapshot: 13.11.2026 · package version 4f099d465ef2 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100salesforce-agentforce
    Use when building AI agents on the Salesforce platform using Agentforce DX, defining agent topics and actions via Salesforce metadata, configuring Einstein Trust Layer, implementing prompt templates, creating Apex-backed agent actions, or deploying Agentforce 360 capabilities…
  • 100agentforce-bot-upgrade
    Use this skill to upgrade Einstein Bots into Agentforce agents end-to-end in a single pass, orchestrating per-bot Agent Spec generation, agentforce-generate authoring, and post-conversion .agent enhancements. TRIGGER when: user asks to migrate, upgrade, or convert one or more…
  • 100agentforce
    Builds, wires, tests, secures and operates Salesforce Agentforce agents in the new Agentforce Builder: Agent Script (.agent), subagents, actions (Flow, Apex, prompt templates, Integration Procedures, External Services, MCP), Prompt Builder, Data Cloud / Data 360 retrievers,…
  • 90agentforce-bot-upgrade
    Use this skill to Upgrade Einstein Bots into Agentforce agents end-to-end in a single pass, orchestrating per-bot Agent Spec generation, planner reconciliation across bots, agentforce-generate authoring, and post-conversion .agent enhancements. TRIGGER when: user asks to…
  • 100sf-einstein
    Implement Einstein AI for Salesforce Commerce — Einstein Recommendations (product-to-product, user-to-product, trending), predictive sort, search ranking, Einstein Search Dictionaries, and Data Cloud personalization. Use when adding AI-powered features to commerce storefronts.
  • 100sales-agentforce-pipeline-management-configure
    Use to configure, set up, or repair the Sales Management agent and Agentforce Pipeline Management in a Salesforce org. Automates metadata creation for flows, prompt templates, permission sets, and data source configuration. TRIGGER when: user wants to enable Pipeline Management,…

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.