What do you guys use for dev environment to protect yourself from Claude Code actually reading any secrets from your computer - protect yourself from leaking any secrets, or have prompt injections?
Commit
9051a0c
Everything runs in a docker in the cloud.
If I do something locally it's extremely limited.
How does claude securely reach the inside of that container, from your machine?
You can see the whole setup here: https://github.com/desplega-ai/agent-swarm/blob/main/src%2Fhttp%2Fmcp.ts
Also you can see how we handle secrets, keys, and other sensitive data paths.
GitHubagent-swarm/src/http/mcp.ts at main · desplega-ai/agent-swarmYour Company Agentic Operating System. Contribute to desplega-ai/agent-swarm development by creating an account on GitHub.
commit 9051a0c