PranavNagrecha/AwesomeSalesforceSkills / skills/admin/einstein-activity-capture-setup
einstein-activity-capture-setup
VerifiedConfigure Einstein Activity Capture (EAC) for an org: enable the feature, create Configuration profiles, set auth type (User-Level or Org-Level OAuth), define email and calendar sync scope, excluded domains and privacy settings, assign users, verify sync health. NOT for querying synced emails or events in SOQL, Apex or reports — use apex/einstein-activity-capture-api. NOT for Opportunity Scoring or Pipeline Inspection — use agentforce/einstein-copilot-for-sales.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 18.6 KB |
| references/examples.md | markdown | 5.7 KB |
| references/gotchas.md | markdown | 5.6 KB |
| references/llm-anti-patterns.md | markdown | 9.1 KB |
| references/well-architected.md | markdown | 6.4 KB |
| scripts/check_einstein_activity_capture_setup.py | script | 11.2 KB |
| templates/einstein-activity-capture-setup-template.md | markdown | 3.4 KB |
references/llm-anti-patterns.md
View on GitHub# LLM Anti-Patterns — Einstein Activity Capture Setup
Common mistakes AI coding assistants make when generating or advising on Einstein Activity Capture Setup.
These patterns help the consuming agent self-check its own output.
## Anti-Pattern 1: Omitting PSL Assignment as a Required Step
**What the LLM generates:** A setup guide that says "Enable EAC in Setup, create a Configuration profile, and add users to the profile." The guide skips or buries the Permission Set License assignment step, treating it as optional or self-evident.
**Why it happens:** LLMs conflate EAC user assignment (adding users to a Configuration profile) with the separate PSL entitlement step. Training data often shows only the Configuration profile UI, not the PSL assignment flow in a different area of Setup.
**Correct pattern:**
```text
REQUIRED sequence for EAC user enablement:
1. Setup > Company Information > Permission Set Licenses
→ Confirm "Einstein Activity Capture" PSL has available seats
2. For EACH user: Setup > Users > [User] > Permission Set License Assignments
→ Assign "Einstein Activity Capture" PSL
→ Bulk assignment via Data Loader INSERT on PermissionSetLicenseAssign is supported
3. ONLY after PSL is assigned:
→ Open EAC Configuration profile
→ Add user to Assigned Users list
If PSL is not assigned first, Configuration assignment appears to succeed
but sync never starts and no error is surfaced.
```
**Detection hint:** Flag any EAC setup guide that does not mention "Permission Set License" or "PSL" as a distinct step before Configuration profile assignment.
---
## Anti-Pattern 2: Recommending Org-Level OAuth for Google Workspace
**What the LLM generates:** "For large rollouts, use Org-Level OAuth to avoid per-user authorization" — applied generically without distinguishing between Microsoft 365 and Google Workspace. The guide instructs admins to configure Org-Level OAuth for Google Workspace users.
**Why it happens:** LLMs describe Org-Level OAuth as the best practice for enterprise rollouts (which is true for Microsoft 365) but fail to apply the platform constraint that Org-Level OAuth is Microsoft 365-only. Google Workspace API architecture does not support the same AAD-style tenant consent model.
**Correct pattern:**
```text
Auth type selection by email platform:
Google Workspace (Gmail / Google Calendar):
→ User-Level OAuth ONLY
→ Each user must individually connect their account in:
Profile > Settings > Connected Accounts
→ No Org-Level OAuth option exists for Google Workspace
Microsoft 365 / Exchange Online:
→ User-Level OAuth: each user connects own account (simpler setup, more support overhead)
→ Org-Level OAuth: admin consents via Azure AD, no per-user action needed
(preferred for rollouts >50 users)
Mixed-platform orgs: create SEPARATE Configuration profiles per email platform
```
**Detection hint:** Flag any guide that recommends Org-Level OAuth without first confirming the email platform is Microsoft 365, or that uses "Org-Level OAuth" and "Google" in the same instruction.
---
## Anti-Pattern 3: Claiming EAC Queryability Without Version Context
**What the LLM generates:** Either "EAC-synced emails are stored as EmailMessage records and you can query them with SOQL" — OR the opposite: "EAC data is in a separate Einstein data store and cannot be queried via SOQL." Either claim may be wrong depending on the org's Salesforce release version.
**Why it happens:** The storage model for EAC changed in Summer '25. Pre-Summer '25 training data shows the separate Einstein data store model (not SOQL queryable via standard objects). Post-Summer '25, synced emails are standard EmailMessage records. LLMs interpolate from whichever version dominates their training data without flagging the version dependency.
**Correct pattern:**
```text
EAC data storage — version-dependent behavior:
Pre-Summer '25:
→ Synced emails: separate Einstein data store, NOT standard EmailMessage records
→ NOT queryable via SOQL on Task, Event, or EmailMessage
→ Visible only in Activity Timeline UI component
→ Does NOT count against org file storage
Summer '25+:
→ Synced emails: standard EmailMessage records, SOQL queryable
→ SELECT Id, Subject, FromAddress, ToAddress FROM EmailMessage WHERE ...
→ DOES count against org file storage
Always confirm the org's Salesforce release version before advising on queryability.
```
**Detection hint:** Any EAC advice about SOQL queryability or storage that does not mention Salesforce release version dependency should be flagged for review.
---
## Anti-Pattern 4: Treating Exclusion Rules as Post-Launch Configuration
**What the LLM generates:** A rollout guide that says: "Enable EAC, assign users, and then add exclusion rules for any domains you don't want to sync." Or an incident response guide that says: "Now that sensitive emails have synced, add the domain to the exclusion list to stop future syncs and the existing records will be cleaned up."
**Why it happens:** LLMs present exclusion rules as a configuration detail rather than a pre-launch gate. They also incorrectly imply that adding an exclusion rule retroactively cleans up already-synced data — mirroring how blocklist systems work in other contexts.
**Correct pattern:**
```text
Exclusion rule behavior:
→ Exclusions are NOT retroactive
→ Adding an exclusion prevents FUTURE syncs from the excluded domain/address
→ Previously synced emails from that domain REMAIN on Salesforce records
→ Manual deletion required if retroactive removal is needed:
SELECT Id FROM EmailMessage WHERE FromAddress LIKE '%excluded-domain.com%'
→ Hard-delete via Data Loader
Pre-launch gate requirement:
→ Obtain signed-off exclusion list from Legal and Compliance BEFORE any users
are added to a Configuration profile
→ Exclusion configuration must be complete and reviewed before first user sync
```
**Detection hint:** Flag any setup guide that describes exclusion rule configuration as a step that comes AFTER user assignment or as a post-launch cleanup mechanism.
---
## Anti-Pattern 5: Ignoring Storage Growth for High-Volume Orgs on Summer '25+
**What the LLM generates:** A capacity plan or storage estimate for EAC that either states EAC does not consume org storage (pre-Summer '25 behavior), or omits any mention of storage growth calculations for high-volume deployments.
**Why it happens:** Until Summer '25, EAC lived outside org storage. Most LLM training data reflects this model. Even for post-Summer '25 knowledge, LLMs often omit volume-based storage calculations because they require combining EAC email volume estimates with per-message storage — a multi-step calculation that LLMs frequently skip.
**Correct pattern:**
```text
EAC storage planning for Summer '25+ orgs:
Estimate monthly storage growth:
(Users) x (avg emails/day) x (avg email size KB) x 30 days / 1,024 = GB/month
Example: 500 reps x 40 emails/day x 20 KB x 30 / 1,024 ≈ 11.7 GB/month
Planning steps:
1. Check current org storage: Setup > Storage Usage
2. Calculate EAC-driven monthly storage growth using the formula above
3. If growth will exceed 80% of storage capacity within 6 months:
a. Purchase additional storage, OR
b. Configure email retention/archival policy, OR
c. Limit EAC sync to high-value users only
4. Set a storage alert at 80% capacity
5. Reassess storage quarterly after go-live
Do NOT enable EAC for all users on Summer '25+ without a storage plan.
```
**Detection hint:** Flag any EAC enablement guide for Summer '25+ orgs that does not include a storage growth calculation or storage planning step for deployments with more than 50 active users.
---
## Anti-Pattern 6: Assuming All Synced Emails Appear on Related Records
**What the LLM generates:** "Once EAC is configured, all synced emails will automatically appear on the Activity Timeline of the related Contact, Lead, Opportunity, and Account records."
**Why it happens:** This is the intended behavior, and LLMs state the ideal outcome without surfacing the matching failure modes. EAC matching depends on email address data quality — a prerequisite that LLMs commonly omit.
**Correct pattern:**
```text
EAC email matching behavior:
→ EAC matches emails to records by comparing sender/recipient email addresses
against Contact and Lead records in the org
→ If NO matching Contact or Lead is found: email goes to Unresolved Items queue
→ If MULTIPLE Contacts share the same email address: match is ambiguous,
email may go to Unresolved Items or match to the first found
→ Matched email appears on Activity Timeline of the matching Contact AND
any related Opportunity/Account/Case via the Contact relationship
Pre-launch data quality steps:
1. Confirm key Contacts have Email field populated
2. Run duplicate Contact report filtered by Email address
3. Resolve duplicate emails before go-live
4. After go-live: monitor Setup > Einstein Activity Capture > Unresolved Items weekly
Never claim "all emails will appear on records" without verifying Contact email data quality.
```
**Detection hint:** Flag any EAC setup guide that does not mention the Unresolved Items queue or Contact email address matching requirements.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deployment | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |