Our projects grow out of everyday work with AI agents. We’re interested in how people and agents share context, work through ideas, and learn from the process.
Human agency is part of that: understanding the work, contributing judgment, and shaping what happens next.
Reopens Claude Code, Codex, Pi, and omp conversations in the panes Rex restores after its server stops, and carries worker results back to the orchestrator.
Stores API keys, tokens, passwords, and SSH keys in iCloud Keychain and gives them to commands as environment variables, masking their values in output.
Checks Claude Code, Codex, and Pi tool calls on macOS before they run. Blocks broad filesystem scans and recognizable credential reads, and tells the agent what to run instead.
Pi extensions for inline skill invocation, session IDs in the footer, Agent Guard checks, and a DeepSeek reasoning workaround. Each can be enabled on its own.
Reopens Claude Code, Codex, Pi, and omp conversations in the panes Rex restores after its server stops, and carries worker results back to the orchestrator.
Checks Claude Code, Codex, and Pi tool calls on macOS before they run. Blocks broad filesystem scans and recognizable credential reads, and tells the agent what to run instead.
Pi extensions for inline skill invocation, session IDs in the footer, Agent Guard checks, and a DeepSeek reasoning workaround. Each can be enabled on its own.
Stores API keys, tokens, passwords, and SSH keys in iCloud Keychain and gives them to commands as environment variables, masking their values in output.
There was an error while loading. Please reload this page.
Most used topics
Loading…
Uh oh!
There was an error while loading. Please reload this page.
You are visitor 01402503 — GeezerVison does not store the pages you look up. Best viewed with Netscape Navigator 4.7 or Internet Explorer 6 at 800×600. No cookies required.