Feature Request - Add the ability to add agents that can be run when using a Claude model.
Currently my team and I do code review and PR creation outside of AI-FDE because we use agents that follow our own patterns and rules. If someone forgets to do these steps outside of the AI-FDE framework it ends up making code review a longer process. It would be nice to include this tooling directly in AI-FDE so everything is consistent.
Hey Shaler
Agent support inside AI-FDE isn’t available yet, but there are a few approaches that can close most of this gap today:
1. Palantir MCP + external agent (best coverage)
If your team uses VS Code or Cursor, you can connect the Palantir MCP server to Claude Code (or any MCP-compatible agent) and run your custom code review / PR agents from there — with full read/write access to your Code Repositories in Foundry. This means your agents operate directly against the same repos AI-FDE touches, keeping everything in sync.
Setup: https://www.palantir.com/docs/foundry/palantir-mcp/overview
2. AI-FDE Skills layer as a lightweight shim
AI-FDE exposes a Skills system where domain skills map to specific tool calls. You can’t inject a full external agent today, but you can structure your review checklist as a Notepad document and reference it as context when prompting AI-FDE. Not ideal, but it enforces the checklist without leaving the tool.
3. AIP Chatbot Studio as a review gate
If your team has AIP access, you can build a dedicated code review chatbot in AIP Chatbot Studio backed by your team’s rules/patterns (as a custom content source). You’d trigger it as a named step after AI-FDE finishes; still two tools, but at least the second one lives inside Foundry and can be made mandatory via a Workshop workflow.
Felipe Montes
Development Advisor
Follow me @ linkedin.com/in/felipemontesl/ or GitHub: Brazillianerd