
from salesforce-claude-code8
Guides building Agentforce AI agents: agent script patterns, actions (Apex invocable), testing workflows, deployment and metadata best practices.
This skill documents procedures and best practices for developing Agentforce AI agents on Salesforce. It covers Agent Script structure, action definitions (Apex @InvocableMethod patterns), testing strategies, CLI workflows for generating and publishing authoring bundles, and deployment metadata ordering. The content serves as a development and CI/CD reference for engineering teams building conversational agents integrated with Salesforce.
Use when authoring or maintaining Agentforce agents, creating Apex invocable actions for agent integration, generating authoring bundles, or designing test suites and CI pipelines for agent validation. Particularly useful for teams using Claude Code, Cursor, or Salesforce tooling.
_reference/ are present).Primarily relevant to developer-facing agents and tooling that integrate with Salesforce (Claude Code, Cursor, or other code-focused agent runtimes).
This skill has not been reviewed by our automated audit pipeline yet.
Search-First (Research Before You Code)
Agent workflow to search for existing Salesforce tools, libraries, MCP servers, and packages before writing new code—helps decide whether to adopt, extend, or b
Salesforce Deployment Constraints
Enforce safe Salesforce deployments: validation-first workflow, test-coverage gates, dependency ordering, pre-deploy snapshots, and rollback readiness.
Sf Docs Lookup
Use when looking up official Salesforce documentation. Apex reference, LWC docs, platform guides, API references, and SF CLI commands via WebSearch.