forcedotcom/sf-skills / plugins/builder/agentforce-adlc/skills/agentforce-generate
agentforce-generate
FlaggedBuild, modify, audit, repair, optimize, debug, and deploy agents with Agentforce Agent Script. TRIGGER when: user creates, reviews, or changes .agent files or aiAuthoringBundle metadata; asks to fix AgentScript, audit an existing agent, run an AgentScript health check, common-pitfall review, or baseline-versus-candidate repair loop; changes a response, action, subagent, route, state flow, or Agent Spec; migrates or upgrades a legacy GenAiPlannerBundle Agentforce agent (NOT an Einstein Bot) to an aiAuthoringBundle (Agent Script) agent; previews, debugs, deploys, publishes, or tests agents; uses sf agent generate/preview/publish/test; or manages Agentforce MCP servers, tools, assets, or authentication. DO NOT TRIGGER when: Apex, Flow, Prompt Template, Experience Cloud, or general Salesforce CLI work is unrelated to Agent Script; the source is an Einstein Bot (BotDefinition/BotVersion) — use agentforce-bot-upgrade; or the primary input is a production session or trace ID rather than an agent artifact.
assets/agents/voice-knowledge-grounded.agent
View on GitHub# Voice + Knowledge-Grounded Agent
# ================================
#
# A voice-channel service agent that answers caller questions by searching an
# Agentforce Data Library (ADL) via the AnswerQuestionsWithKnowledge action.
# This is the Project Codey "Steel Thread 2" shape: voice + knowledge grounding.
#
# Combines:
# - modality voice: (platform default voice — customize later in Agent Builder)
# - VoiceCallId linked variable + connection messaging/customer_web_client
# - knowledge: block + AnswerQuestionsWithKnowledge action
#
# Replace ARFPC_<libraryId> below with the value computed from your provisioned
# ADL: rag_feature_config_id = "ARFPC_" + libraryId.
# See references/data-library-reference.md for provisioning.
system:
instructions: "You are a voice-based AI customer service agent. Keep responses brief — callers are listening, not reading. Answer questions ONLY from the knowledge base; never guess or read out information you did not retrieve. Never read out URLs, citations, or visual formatting — summarize what matters instead. Read back critical data before acting on it."
messages:
welcome: "Hi, thanks for calling! What can I help you with today?"
error: "I'm sorry, I'm having trouble with that right now. Could you say that again?"
access:
default_agent_user: "agent@example.com"
config:
developer_name: "VoiceKnowledgeAgent"
agent_label: "Voice Knowledge Agent"
description: "Voice-channel service agent that answers caller questions grounded on an Agentforce Data Library"
agent_type: "AgentforceServiceAgent"
variables:
EndUserId: linked string
source: @MessagingSession.MessagingEndUserId
description: "End user identifier"
RoutableId: linked string
source: @MessagingSession.Id
description: "Routable session identifier"
ContactId: linked string
source: @MessagingEndUser.ContactId
description: "Contact record identifier"
VoiceCallId: linked string
source: @VoiceCall.Id
description: "This variable may also be referred to as Voice Call Id"
knowledge:
rag_feature_config_id: "ARFPC_<libraryId>"
citations_enabled: True
citations_url: ""
language:
default_locale: "en_US"
additional_locales: ""
all_additional_locales: False
# Minimum voice config: platform default voice ("Mark", en_US) + core tuning.
# Optional advanced settings (filler-word detection, speak-up prompts, endpointing)
# are documented in references/voice-modality-reference.md — add them only if needed.
# Customize the voice later in Agent Builder → Connections → Voice.
modality voice:
voice_id: "UgBBYS2sOqTuMpoF3BR0"
outbound_speed: 1
outbound_stability: 0.65
outbound_similarity: 0.75
# customer_web_client (ECv2) is the voice-capable surface. connection messaging is
# additive here — kept because this agent escalates to a human (@utils.escalate).
connection customer_web_client:
adaptive_response_allowed: True
connection messaging:
escalation_message: "Let me transfer you to a specialist who can help with that."
start_agent agent_router:
description: "Welcomes the caller and routes to the appropriate topic"
reasoning:
instructions: ->
| Listen to what the caller needs and route them to the right topic.
For questions about products, policies, procedures, or troubleshooting, go to knowledge answers.
If the caller asks to speak with a person or you cannot help, escalate.
Do NOT answer questions directly — route to the appropriate topic.
actions:
go_knowledge: @utils.transition to @subagent.knowledge_answers
description: "Route to knowledge answers for product, policy, or how-to questions"
go_escalation: @utils.transition to @subagent.escalation
description: "Transfer to a human agent"
subagent knowledge_answers:
description: "Answers caller questions by searching the knowledge base"
reasoning:
instructions: ->
| If the question is unclear, ask one short clarifying question first.
For every substantive question, call AnswerQuestionsWithKnowledge
before answering.
Answer only from the returned knowledge summary, using one or two
spoken sentences.
If the summary is empty, say the knowledge base has no answer and
use {!@actions.go_escalation}.
Never speak URLs, citations, or list formatting. Offer to send
returned links instead.
After answering, ask if there's anything else you can help with.
actions:
AnswerQuestionsWithKnowledge: @actions.AnswerQuestionsWithKnowledge
with query = ...
with citationsUrl = ...
with ragFeatureConfigId = ...
with citationsEnabled = ...
go_escalation: @utils.transition to @subagent.escalation
description: "Transfer to a human agent when the knowledge base has no answer or the caller asks for a person"
go_back: @utils.transition to @subagent.agent_router
description: "Return to main routing when the caller has a different question"
actions:
AnswerQuestionsWithKnowledge:
description: "Answers questions about company policies, procedures, troubleshooting, or product information by searching knowledge articles. For example: 'What is your return policy?' or 'How do I fix an issue?'"
inputs:
query: string
description: "Required. A string created by generative AI to be used in the knowledge article search."
label: "Query"
is_required: True
is_user_input: True
citationsUrl: string = @knowledge.citations_url
description: "The URL to use for citations for custom Agents."
label: "Citations Url"
is_required: False
is_user_input: True
ragFeatureConfigId: string = @knowledge.rag_feature_config_id
description: "The RAG Feature ID to use for grounding this copilot action invocation."
label: "RAG Feature Configuration Id"
is_required: False
is_user_input: True
citationsEnabled: boolean = @knowledge.citations_enabled
description: "Whether or not citations are enabled."
label: "Citations Enabled"
is_required: False
is_user_input: True
outputs:
knowledgeSummary: object
description: "A string formatted as rich text that includes a summary of the information retrieved from the knowledge articles and citations to those articles."
label: "Knowledge Summary"
complex_data_type_name: "lightning__richTextType"
filter_from_agent: False
is_displayable: True
citationSources: object
description: "Source links for the chunks in the hydrated prompt that's used by the planner service."
label: "Citation Sources"
complex_data_type_name: "@apexClassType/AiCopilot__GenAiCitationInput"
filter_from_agent: False
is_displayable: False
target: "standardInvocableAction://streamKnowledgeSearch"
label: "Answer Questions with Knowledge"
require_user_confirmation: False
include_in_progress_indicator: True
progress_indicator_message: "Let me check on that for you"
source: "EmployeeCopilot__AnswerQuestionsWithKnowledge"
subagent escalation:
description: "Transfers the caller to a live human agent"
reasoning:
instructions: ->
| The caller needs to speak with a human agent. Use {!@actions.escalate_to_human} to transfer them.
Before transferring, briefly let them know you're connecting them.
actions:
escalate_to_human: @utils.escalate
description: "Transfer the caller to a live agent"
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |