Create a reusable Copilot profile that helps with LoadRunner and VuGen questions by adding an agent Markdown file to the location supported by your chosen Copilot host. The profile can guide explanations and draft script-related suggestions; it does not install VuGen, connect to LoadRunner, or prove that a script or load test will run.
What a Copilot custom agent does
A custom agent is a reusable, role-specific configuration: its profile identifies the agent, describes its capabilities and tool access, and provides behavioral instructions. For GitHub Copilot cloud agent, GitHub documents a Markdown file with YAML frontmatter followed by a Markdown instruction body. The frontmatter can include fields such as description, name, target, tools, model, user-invocable, disable-model-invocation, mcp-servers and metadata. Which fields and behaviors apply depends on the host. See GitHub’s custom agents configuration reference.
For this example, the agent’s role is to help developers understand or draft LoadRunner/VuGen-related virtual-user script changes using the context they provide. SAP’s 2023 Virtual User Generator Installation Guide describes VuGen as a tool for creating virtual-user (Vuser) scripts, primarily through recording. That context supports a focused instruction profile, but it does not establish a dedicated Copilot–LoadRunner integration.
Choose where the agent will run
Choose the host before writing the profile: file locations, supported frontmatter, available tools and selection steps differ between GitHub.com cloud agent and IDE agents.
#1 Best Overall
| Use case | Profile location | What to know |
|---|---|---|
| GitHub Copilot cloud agent for a repository | .github/agents |
GitHub’s creation guide uses filenames such as my-agent.agent.md. Commit and merge the profile into the repository’s default branch before it appears in that repository’s agent picker. |
| VS Code agent shared with a workspace | .github/agents |
Microsoft documents workspace agents here; available controls depend on the selected agent harness. |
| VS Code agent for an individual user | ~/.copilot/agents or ~/.claude/agents |
Use a user-level location when the profile should be reused across that user’s workspaces rather than shared through a repository. |
GitHub’s cloud-agent creation guide covers repository profiles and the cloud workflow: Creating custom agents for Copilot cloud agent. For VS Code file locations and creation options, see Microsoft’s Custom agents in Visual Studio Code.
Create the profile
For a GitHub cloud-agent repository profile, create a Markdown file under .github/agents. The sample below is an editorial starter, not a vendor-provided LoadRunner template. It explicitly requests only read and search access as an example; confirm the tool names supported by your host and repository before relying on them.
---
name: loadrunner-helper
description: Helps explain and draft LoadRunner/VuGen virtual-user script work from repository context.
tools: [read, search]
---
You help developers understand and draft LoadRunner/VuGen-related virtual-user script changes.
Before proposing code, ask for the LoadRunner version, protocol, and relevant script or error details when they are missing. Explain assumptions, keep suggestions grounded in supplied repository context, and distinguish illustrative code from verified vendor syntax. Do not claim to execute VuGen or run a load test unless the user has provided an available tool and evidence that it ran. State what a developer should review in the installed version's documentation before relying on version-specific behavior.The description is required in GitHub’s reference. Add a friendly name if useful in the picker. GitHub says that omitting tools enables all available tools; an explicit list can constrain access, and tools: [] disables tools. MCP tools may be named or namespaced. Grant only what the role needs, and verify names available in the selected environment.
Do not copy fields across hosts on the assumption that they are universal. GitHub’s cloud-agent reference says IDE custom-agent fields such as argument-hint and handoffs are not supported for Copilot cloud agent on GitHub.com. Check the target’s current reference for the fields you intend to use.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Write instructions that make the agent useful
The frontmatter describes the profile to the host; the body explains how it should work. For a LoadRunner/VuGen helper, concrete instructions can ask for the missing context before suggesting code and make uncertainty visible.
- Request relevant inputs: LoadRunner version, protocol, the relevant script excerpt, and the error or task. Ask follow-up questions when these details are needed to avoid guessing.
- Define the expected output: explain existing script behavior, suggest a focused change, or draft illustrative code, as appropriate to the request.
- Set an evidence boundary: distinguish an example from verified vendor syntax; tell the developer which version-specific behavior to check in the help for their installed version.
- Set a tool boundary: do not imply that the profile itself supplies VuGen, connects to a running LoadRunner installation, executes a script, or runs a load test.
These are design choices for a helpful profile, not claims about a tested integration or current LoadRunner feature set. SAP’s guide is a 2023 edition; consult the relevant help documentation for the LoadRunner version actually in use rather than treating that guide as current version-specific advice.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Choose between direct editing and the VS Code editor
| Approach | Best suited to | Trade-off |
|---|---|---|
| Edit the Markdown file directly | People who want direct control over the file, frontmatter and instruction body. | You must ensure the fields and syntax match the intended host. |
| Use VS Code’s Agent Customizations editor | People who want to create a profile through the VS Code interface. | Review the generated frontmatter, tool choices and instructions for accuracy and for the selected harness. |
Microsoft documents both manual creation and creation through the Agent Customizations editor. Whichever route you use, treat the resulting profile as a configuration to inspect, not an automatic guarantee of correct tool access or behavior.
Select the profile and review its behavior
- For GitHub cloud agent: commit the profile and merge it into the repository’s default branch. After that, select it from that repository’s agent picker. GitHub also documents selecting a custom agent in Copilot CLI with
/agent, a prompt reference, or a command-line argument. - For VS Code: select the intended harness and agent in the Chat or Agents interface. The available controls depend on the harness.
- Review representative responses and tool use: try a question with version and protocol details, then one with essential details missing. Check that the agent asks for context when needed, labels unverified syntax, and does not claim to have executed VuGen or a test without evidence.
GitHub’s cloud-agent workflow is documented in its custom-agent creation guide; VS Code’s steps and review guidance are in Microsoft’s custom-agent documentation.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
