Integration with RChilli MCP

The RChilli MCP Server integration with Salesforce Agentforce enables customers to use RChilli capabilities directly within their Agentforce workflows.

By making RChilli MCP tools available to Agentforce agents, organizations can automate recruitment-related tasks and allow agents to retrieve, process, and use candidate or resume information as part of a conversational workflow

For example, a recruiting agent can use an RChilli MCP tool to parse a candidate resume, extract relevant information, and use the returned data to respond to recruiter queries or support downstream recruiting activities.

This integration helps customers extend Salesforce Agentforce with RChilli intelligence while keeping the user experience within Salesforce.

  1. Sign in to Salesforce and go to Setup.

  2. In Quick Find, search for Registered MCP Servers, and then select Registered MCP Servers.

  3. On Agentforce Registry page, select New, and then select Register MCP Server.

  4. On the Create Connection pop-up, enter the RChilli MCP Server details and click Create and Continue. Refer below table for more details.

    Field Value
    MCP Server Name RChilliMCP
    Description 
    RChilli MCP provides access to resume, job, taxonomy, and
                      other RChilli capabilities.
    Server URL https://mcp.rchilli.ai/mcp
    Authentication Method OAuth 2.0
    Identity Provider URL https://mcp.rchilli.ai/token
    Scope As provided by RChilli, contact support@rchilli.com.
    Client ID Your RChilli client ID, contact support@rchilli.com.
    Client Secret Your RChilli client secret, contact support@rchilli.com.
  5. After the connection is created, click on the RChilli MCP you created.

  6. On the MCP Server Allowlist pop-up, select the tools that the Agentforce Agent needs.
    • Review each tool name and description
    • Select the required tools
    • Review any Salesforce MCP risk assessment shown for the server/tools
    • Once tools are selected, click Allow and Continue.

  7. On the Apply Policies pop-up, click Save.

  8. In Quick Find, search for Agentforce Studio, and then select Agentforce Agents. On the Agentforce Agents page:
    • Click the Agentforce toggle button to ON.
    • Click + New Agent to create a new agent.

Create Agent

  1. On the Agentforce Builder, enter the purpose of the agent, such as Recruiter Assistant, and click on the continue symbol.

  2. Enter the description of the agent and click on the continue symbol.

  3. On the Name your agent page:
    • Enter Agent Name
    • You can click on the Select User if you want to assign an existing User to manage the agent access or you can click on the New User.
    • Click Let's Go

  4. In the Subagents section, select the + icon, and then select New Subagent.

  5. Enter the Subagent Name and Description.
    For example:
    • Subagent Name: Job Assistant Agent
    • Description:
      • Generate job descriptions.
      • Parse job descriptions.
      • Generate interview questions.
      • Search the skill taxonomy.
      • Search the job profile taxonomy.
    • Select Create and Open.

  6. In the newly created subagent, select the + icon, and then select Add from Asset Library.

  7. Search RChilli and then select the RChilli MCP tools required for the subagent, and then select Add to Agent.

  8. Verify that the selected RChilli MCP tools appear under the subagent and click Preview to preview the agent.

  9. Enter a request that should use one of the configured RChilli MCP tools. Agent will retrieve the information based on the MCP tools you added and based on the prompts.

    For example:

    Give me skills related to
      Java.


Note:

You can create additional subagents based on your recruitment use cases. To create another subagent, repeat the step-4 to step-9 and add only the RChilli MCP tools required for that use case.

For example, you can create subagents for:

  • Candidate Matching — Compare a candidate resume with a job description and evaluate the candidate’s overall fit.
  • Interview Question Generation — Generate role-specific interview questions based on the job description or candidate profile.
  • Job Description Processing — Parse job descriptions and identify key requirements, skills, and qualifications.
  • Skill and Job Profile Search — Search RChilli skill and job profile taxonomy data.
  • Candidate Screening — Evaluate a candidate against predefined job requirements and identify strengths or gaps.

Create each subagent for a specific purpose and add only the MCP tools that support that workflow. This helps keep the agent configuration focused and ensures that the appropriate RChilli MCP tools are used for each task.