forcedotcom/sf-skills/sales-call-scoring-configure

Claude Code and agent skill by forcedotcom/sf-skills

Trust tier 2
  • community
LLM review pending
Source check pending

Checking that the SKILL.md file is still in the repository.

Configure, set up, or repair AI Call Scoring (a.k.a. Call Coaching / Coaching Competencies) on an ECI-enabled Salesforce org. Flips ECICallScoringEnabled via SOAP Metadata API, enables Einstein Generative AI Platform if off, and manages Coaching Competency records (EnablementCompetencyDef Tooling API): install the 6 OOTB set, the 7 Momentum best-practice prompts, or conversationally author/edit/activate/deactivate custom competencies (8-active + 16000/255-char caps). TRIGGER when: user wants to turn on AI Call Scoring or Call Coaching, install OOTB or best-practice competencies, create/edit custom competencies, re-enable after sandbox refresh, or verify/repair partial Call Coaching setup on a 68+ ECI org. DO NOT TRIGGER when: no ECI license (direct to Agentforce for Sales), user wants to enable ECI/Momentum themselves (out of scope — Setup), Pipeline Management (use sales-agentforce-pipeline-management-configure), or custom agent (use agentforce-generate).

Install

npx vskill@latest install forcedotcom/sf-skills/sales-call-scoring-configure

SKILL.md

The same SKILL.md was published earlier as forcedotcom/sf-skills/sales-call-scoring-configure.

Sales Call Scoring — Configure

Configure Call Coaching (a.k.a. AI Call Scoring / Coaching Competencies) on a Salesforce org. Detects the org's current state, enables dependent features it is authorized to enable, and flips the ECICallScoringEnabled org preference via the SOAP Metadata API.

Scope

In scope

  • ECI-enabled orgs only. The skill supports orgs on the ECI (Einstein

Conversation Insights) arm (spike scope: ECI-only for MVP).

  • Detect which of the 4 PRD-scoped org states the org is in.
  • Enable Einstein Generative AI Platform if off (PRD-authorized dependency).
  • Flip enableECICallScoring to true via SOAP Metadata API v68.
  • Verify the change took effect.
  • Idempotent — clean exit on states 3 and 4 (already enabled).

Out of scope

  • River Rush / Momentum detection. Deferred until that arm goes live —

see Key Considerations. A Momentum-only org today is not distinguished from any other ECI-off org; it gets the generic "ECI is not enabled" message from Phase 3, not a Momentum-specific one.

  • Enabling ECI or Momentum themselves. PRD-explicit out-of-scope; the skill

detects and messages the admin to enable ECI via Setup.

  • License provisioning. Missing base licenses (Agentforce for Sales, ECI

license) require an account-team ticket.

  • Competency CRUD. Handled by sibling scripts in this same skill — see "Competency CRUD workflow" below (§S3-§S10) and the "Execution sequence" matrix.

Prerequisites

  1. `sf` CLI authenticated — sf org login web --alias <org-alias>.
  2. ECI arm licensed and enabled: ConversationPilot org permission (Phase 0

purchase gate) and enableCallCoaching pref (Phase 1, the ConversationPilotPref preference). (River Rush/Momentum is a separate arm covered by the same server-side gate but not checked by this skill yet — see Key Considerations.)

  1. Site gate com.salesforce.eci.callScoringEnabled open on the org's pod

(RelEng-controlled; not admin-toggleable).

Einstein GenAI Platform is a soft prerequisite — the skill enables it automatically when off.

Clarifying Questions

Ask the admin: target org alias? Sandbox refresh scenario? (The pref has disableOnSandboxCopy=true, so it must be re-flipped after every refresh — that's normal, not a failure.)

Admin Communication Guidelines

CRITICAL: This skill serves admin users, not developers. The full 13-rule spec with rationale for each rule lives in `references/admin-communication.md` — consult it before your first response in this skill. Summary of the load-bearing rules:

  1. Run all bash commands in background (run_in_background: true).
  2. Speak admin language, not entity language — "Call Coaching", "Coaching Competency", "active/inactive". Never EnablementCompetencyDef, enableECICallScoring, EvaluationInstructions, SourceCompetencyTemplate in primary chat text.
  3. Hide technical details — no record Ids, DeveloperNames, Template enum keys, HTTP codes, SOAP responses, curl/jq internals, script paths, or phase numbers. Exception: the S9 verify markdown table (see `references/competency-crud.md` §S9).
  4. Chain automatically — when a script prints Next: <script>, run it in the same turn without asking. Exception: State 3 → rule #13.
  5. Ask OOTB-vs-custom once — on first-time competency setup (State 3, unqualified intent), render install-ootb-competencies.sh --list and let the admin choose. Only checkpoint in this skill where Claude asks before writing.
  6. Idempotency is silent — if already on/installed, say so briefly and move on.

See `references/admin-communication.md` for the full 13-rule spec, including error-surfacing, outcome-vs-steps reporting, and the S9 carve-out rationale.

Workflow

Phase 0 — Auth

Resolves org auth via the shared scripts/shared/auth.sh helper (resolve_org_auth "$ORG_ALIAS"), which every script in this skill sources: validate the alias format, call sf org display for instanceUrl and accessToken (falling back to sf org auth show-access-token when the CLI redacts the token), refuse a non-HTTPS instance URL, and fail closed with an explicit sf org login web --alias <alias> instruction if the org isn't reachable or no token can be extracted. If ACCESS_TOKEN/INSTANCE_URL are already exported for the same alias — e.g. a parent shell already resolved auth and is chaining multiple of these scripts in one session — resolution is skipped and the cached values are reused. orgId is not part of this helper; enable-call-scoring.sh fetches it separately (via its own sf org display) purely for the status banner.

Phase 0 — ECI purchase gate

Before any state detection, query the ConversationPilot OrgPermission via Tooling API (OrganizationSettingsDetail, same pattern as the ECICallScoringAI gate in Phase 3). This is a license/purchase check, distinct from enableCallCoaching below (the admin's Setup toggle for an org that's already licensed):

  • Not enabled: exit with: "Call Coaching couldn't be turned on because

this org doesn't have Einstein Conversation Insights, which is required to use Call Coaching. To proceed, an admin can go to Setup → Salesforce Go → Einstein Conversation Insights, or contact their sales representative, to understand how to purchase ECI." Non-zero exit.

  • Enabled: proceed to Phase 1's enableCallCoaching check.
Phase 1 — State detection

SOAP readMetadata at /services/Soap/m/68.0 against two settings types:

  1. ConversationalIntelligenceSettings → read enableCallCoaching (ECI arm)

and enableECICallScoring (current Call Coaching state).

  1. EinsteinGptSettings → read enableEinsteinGptPlatform.

(RiverRushSettings/riverRushEnabled — Momentum — is NOT read. That detection is deferred until the River Rush arm goes live; see Key Considerations for where to re-add it.)

Query existing competencies via Tooling API: SELECT COUNT() FROM EnablementCompetencyDef.

Classify the org into one of 4 states:

StateDetection
1a. Not available (no ECI license)Cannot read ConversationalIntelligenceSettings
2. Available, not enabledenableCallCoaching=true AND enableECICallScoring=false
3. Enabled, no competenciesenableECICallScoring=true AND count = 0
4. Enabled, with competenciesenableECICallScoring=true AND count > 0
Phase 2 — Route by state
  • State 1a: Exit with PRD-verbatim message: "AI call coaching is not

available in this org and they need Agentforce for Sales." Non-zero exit.

  • State 3 / State 4: Exit 0 with a ready-for-competency-setup message.
  • State 2: Proceed.
Phase 3 — Prerequisite enablement (State 2 only)
  • License gate (checked first): query OrganizationSettingsDetail

(Tooling API) for the ECICallScoringAI OrgPermission. Only an explicit `false` refuses — exit with: "AI call coaching is not available in this org and org requires Agentforce for Sales license." (Non-zero exit.) If the permission can't be read — the query is rejected with INVALID_TYPE (some org editions, for example Developer Edition) or returns no row — treat it as unknown and proceed; do NOT fail closed on an unreadable signal. The Einstein GenAI TYPE_UNAVAILABLE pre-check catches most unlicensed orgs; the enable updateMetadata write is the final check. This mirrors the Phase 0 ConversationPilot gate, which defers on any non-conclusive read. This runs even when enableCallCoaching/enableECICallScoring already read favorably, since those settings don't by themselves confirm the org is licensed. Must run before the ECI-off check below — a genuinely unlicensed org and a licensed-but-toggled-off org both read enableCallCoaching=false, so checking ECI first would misdiagnose a licensing gap as a self-serve Setup toggle (the regression the NoLicense eval dataset guards against).

  • If ECI is off: refuse with a Setup pointer, exit non-zero. This skill

does NOT enable ECI (PRD explicit). (A Momentum-only org lands here too, since Momentum isn't detected separately — it gets this same generic message, not a Momentum-specific one.)

  • If Einstein GenAI off: SOAP updateMetadata on EinsteinGptSettings

with enableEinsteinGptPlatform=true. PRD explicitly authorizes this. If it's already on, skip the updateMetadata call and tell the admin it's already enabled. If the update fails because the org isn't licensed for EinsteinGptSettings (or is explicitly refused with no fault): tell the admin they need Agentforce for Sales to enable AI Call Scoring, and that without it they can still set up manual call scoring via Salesforce Go setup. A distinct auth/permission or network/timeout failure gets its own message (check CLI session / retry) rather than the licensing message, since those aren't licensing gaps.

Phase 4 — Enable Call Coaching (State 2 only)

If enableECICallScoring already reads true (e.g. it changed out-of-band between Phase 1's read and here), skip the update and tell the admin it's already enabled. Otherwise SOAP updateMetadata on ConversationalIntelligenceSettings with enableECICallScoring=true. Require <success>true</success>.

Phase 5 — Verify (State 2 only)

Re-read enableECICallScoring. Require true. Print SOAP response on failure.

Key Considerations

This is the start of the file. Read the full SKILL.md on GitHub.

Security

  • Security scanPassed, 97 of 100Pattern checks on the skill's files.
  • Model reviewNot runA model review is scheduled for the next rescan.

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

Repository stars
1,054
Repository forks
353
7-day trend score
40

Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/forcedotcom/sf-skills/sales-call-scoring-configure/badge)](https://verified-skill.com/skills/forcedotcom/sf-skills/sales-call-scoring-configure)

Something wrong with this skill or its page? Tell us.