afv-library/skills/developing-agentforce/assets/components/escalation-setup.agent
Willie Ruemmele 261abd679a
chore: rename topic to subagent for Agent Script v2 @W-21955450@ (#193)
* @W-21955450@ Rename topic to subagent for Agent Script v2

Aligns with Agent Script v2 naming standards where `topic` is renamed
to `subagent` across all skill documentation and templates.

Changes:
- Agent Script templates: topic keyword → subagent keyword
- References: @topic.* → @subagent.*
- Documentation: Updated all skill references and guides
- Natural language references preserved in comments/descriptions

* Rename start_agent topic_selector to agent_router

Completes the topic → subagent terminology alignment by:

1. Renaming start_agent from topic_selector to agent_router (15 agent files)
2. Updating template topic declarations: topic {{placeholder}} → subagent {{placeholder}} (5 files)
3. Updating all @subagent.topic_selector references to @subagent.agent_router (35 occurrences)
4. Updating documentation: prose, examples, and diagrams (10 markdown files)
5. Updating comments to use agent_router terminology

Files affected:
- 22 agent template files
- 10 documentation/reference markdown files
- Template component files

The agent_router name is more descriptive of its actual function
(routing to different subagents) and completes the Agent Script v2
terminology standardization.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* Rename files with "topic" to use "subagent" terminology

Completes the topic → subagent terminology alignment by renaming
files and updating all references:

**Files renamed (5):**
- multi-topic.agent → multi-subagent.agent
- template-single-topic.agent → template-single-subagent.agent
- template-multi-topic.agent → template-multi-subagent.agent
- topic-with-actions.agent → subagent-with-actions.agent
- agent-topic-map-diagrams.md → agent-subagent-map-diagrams.md

**References updated (6 docs):**
- Updated all filename references to point to new filenames
- Updated "Topic Map" → "Subagent Map" throughout documentation
- Updated "multi-topic"/"single-topic" → "multi-subagent"/"single-subagent"

Files modified:
- README.md, SKILL.md, agent-spec-template.md
- assets/agents/README.md, assets/README-legacy.md
- references/agent-design-and-spec-creation.md

This ensures consistent "subagent" terminology across filenames,
file content, and all documentation references.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* Complete topic-to-subagent terminology update across skills

Comprehensive update replacing "topic" with "subagent" terminology throughout
the developing-agentforce and testing-agentforce skills to align with Agent
Script's `subagent` block naming.

Key changes:
- "Topic Selector" → "Subagent Router" in all agent templates and docs
- "Topic/action" → "Subagent/action" in documentation
- "Topic map" → "Subagent map" in diagram references
- Updated all architecture documentation to use "subagent" terminology
- Updated 19 .agent template files with new labels and comments
- Updated 8 reference documentation files with consistent terminology

API contract preservation:
- Test spec YAML files preserve "topic" terminology to match Testing Center API
- Added clarifying comments explaining topic/subagent equivalence in YAML files
- Field names like `expectedTopic` unchanged (Salesforce API requirement)

Preserved terms:
- "off-topic" (standard phrase for out-of-scope)
- "expectedTopic" field (Testing Center API)
- "platform topics" (Salesforce guardrail features)

32 files changed, 379 insertions(+), 366 deletions(-)

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* Complete comprehensive topic-to-subagent terminology update

Thorough update replacing all remaining "topic" references with "subagent"
terminology across developing-agentforce, testing-agentforce, and
observing-agentforce skills to fully align with Agent Script's `subagent`
block naming.

Key changes:
- Agent Script syntax: @topic.<name> → @subagent.<name>
- Agent Script syntax: topic.actions → subagent.actions
- Shell script patterns: ^topic → ^subagent
- Documentation: "topic instructions" → "subagent instructions"
- observing-agentforce skill: Updated all agent architecture references
- Template files: Updated all inline comments and descriptions
- Variable names in scripts: TOPIC → SUBAGENT

Specific updates:
- 45 files changed, 294 insertions, 294 deletions
- Updated all Agent Script code examples to use @subagent syntax
- Updated observing-agentforce issue classification guide
- Updated shell script patterns in diagnostic tools
- Updated Apex comments to clarify topic field maps to subagents

Preserved (as required):
- "off-topic" and "off_topic" (standard out-of-scope phrase)
- Testing Center API fields: expectedTopic, topic: in YAML
- API response fields: .topic, generatedData.topic, topic_assertion
- STDM field names: ssot__TopicApiName__c (with clarifying docs)
- Template placeholders in test specs (API values)
- "Topic hash drift" (API field behavior)
- "Email topic/purpose" (means email subject)
- Explanatory comments about API field mapping

All Agent Script syntax and documentation now consistently uses "subagent"
while preserving backward compatibility with platform API field names.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* a few more topic -> subagent replacements

---------

Co-authored-by: Steve Hetzel <shetzel@salesforce.com>
Co-authored-by: Claude Sonnet 4.5 <noreply@anthropic.com>
2026-04-27 12:42:18 -06:00

170 lines
6.3 KiB
Plaintext

# Escalation Setup Pattern
# Complete agent template with connection block for human escalation
#
# ★ When To Use This Pattern:
# - Agent needs to transfer conversations to human agents
# - Using Omni-Channel for routing
# - Enhanced Chat or other messaging channels
#
# ★ Key Components:
# 1. connection messaging: block - defines routing destination
# 2. @utils.escalate action - triggers the transfer
# 3. escalation subagent - handles the handoff flow
#
# ★ Prerequisites:
# - Omni-Channel configured in Salesforce
# - Queue/Skill created for routing
# - Messaging channel active (Enhanced Chat, etc.)
#
# This is a COMPLETE template - customize for your use case
system:
instructions: "You are a helpful customer service agent. Be professional, friendly, and helpful. If you cannot resolve an issue or the customer requests a human, transfer them to a live agent."
messages:
welcome: "Hello! I'm here to help you today. What can I assist you with?"
error: "I apologize, but I encountered an issue. Let me connect you with a human agent."
config:
agent_name: "{{AGENT_NAME}}"
default_agent_user: "{{AGENT_USER_EMAIL}}"
agent_label: "{{AGENT_LABEL}}"
description: "Customer service agent with human escalation capability"
variables:
# Required linked variables for messaging context
EndUserId: linked string
source: @MessagingSession.MessagingEndUserId
description: "Messaging End User ID"
RoutableId: linked string
source: @MessagingSession.Id
description: "Messaging Session ID"
ContactId: linked string
source: @MessagingEndUser.ContactId
description: "Contact ID"
# Escalation tracking variables
escalation_requested: mutable boolean = False
description: "Whether customer requested human agent"
escalation_reason: mutable string
description: "Reason for escalation subagent"
# ⚠️ Use 'number' not 'integer' - integer type is NOT supported in AiAuthoringBundle
attempts_before_escalation: mutable number = 0
description: "Number of attempts before escalating"
language:
default_locale: "en_US"
additional_locales: ""
all_additional_locales: False
# ★ CONNECTION BLOCK - Required for @utils.escalate to work
# This defines where escalated conversations are routed
# Use singular 'connection' for one channel, plural 'connections' for multiple
connection messaging:
# ⚠️ IMPORTANT: Only "OmniChannelFlow" is supported (not "queue", "skill", or "agent")
outbound_route_type: "OmniChannelFlow"
# API name of your Omni-Channel Flow
outbound_route_name: "{{OMNI_CHANNEL_FLOW_NAME}}"
# ⚠️ REQUIRED: escalation_message must be included when connection block is present
escalation_message: "Transferring you to a human agent now..."
# Optional: Allow agent to adapt responses during escalation
adaptive_response_allowed: True
# ★ MULTI-CHANNEL EXAMPLE (use 'connections' plural for multiple channels)
# connections:
# messaging:
# outbound_route_type: "OmniChannelFlow"
# outbound_route_name: "Chat_Support_Flow"
# escalation_message: "Connecting you to chat support..."
# adaptive_response_allowed: True
# telephony:
# outbound_route_type: "OmniChannelFlow"
# outbound_route_name: "Phone_Support_Flow"
# escalation_message: "Transferring to phone support..."
# adaptive_response_allowed: False
# Entry point
start_agent agent_router:
label: "Subagent Router"
description: "Routes users to appropriate subagents based on intent"
reasoning:
instructions: ->
| Greet the customer and determine their needs.
| If they ask for a human or live agent, route to escalation.
| Otherwise, try to help them directly.
actions:
go_help: @utils.transition to @subagent.help
go_escalation: @utils.transition to @subagent.escalation
available when @variables.escalation_requested == True
subagent help:
label: "Help"
description: "Provides assistance to customers"
reasoning:
instructions: ->
| Help the customer with their question.
| If you cannot resolve their issue after 2-3 attempts, offer to connect them with a human.
| If they explicitly ask for a human agent at any time, transfer immediately.
|
| Phrases that indicate escalation request:
| - "talk to a human"
| - "speak to someone"
| - "real person"
| - "live agent"
| - "customer service representative"
|
| Track escalation reason if provided:
set @variables.escalation_reason = ...
actions:
offer_escalation: @utils.transition to @subagent.escalation
immediate_escalation: @utils.transition to @subagent.escalation
subagent escalation:
label: "Escalation"
description: "Transfers conversation to human agent"
reasoning:
instructions: ->
| The customer is being transferred to a human agent.
| Acknowledge their request and apologize for any inconvenience.
| Let them know a human will be with them shortly.
|
| Say something like:
| "I understand you'd like to speak with a human agent. I'm connecting you now.
| A customer service representative will be with you shortly. Thank you for your patience."
actions:
# ★ ESCALATION ACTION
# This transfers the conversation to the queue defined in connection block
transfer_to_human: @utils.escalate
description: "Transfer to human agent when customer requests or issue cannot be resolved"
# ★ NOTE: Skill-Based and Queue-Based Routing
# ⚠️ As of Dec 2025, only "OmniChannelFlow" is supported for outbound_route_type
# "queue", "skill", and "agent" cause validation errors
# You must create an Omni-Channel Flow that routes to your desired queue/skill
# ★ Alternative: GenAiPlannerBundle Escalation with Reason
# If using GenAiPlannerBundle (not visible in Studio), you can use:
#
# actions:
# escalate_with_reason: @utils.escalate with reason="Customer requested human assistance"
#
# NOTE: The "with reason" syntax only works in GenAiPlannerBundle!
# AiAuthoringBundle will fail with SyntaxError if you use it.
# ★ Troubleshooting Escalation:
#
# Issue: "escalate" action not recognized
# Fix: Add the connection messaging: block
#
# Issue: Transfer fails silently
# Fix: Verify Omni-Channel queue exists and has available agents
#
# Issue: SyntaxError: Unexpected 'with'
# Fix: You're using AiAuthoringBundle - remove "with reason" syntax
#
# Issue: Agent user lacks permissions
# Fix: Grant Omni-Channel permissions to the default_agent_user