OpenAI's Codex for Mac adds Chronicle feature using screen captures to enhance AI context
OpenAI released Chronicle for Codex on Mac, a feature that captures screen content to build contextual memories for the AI coding assistant. Available to Pro subscribers as a research preview, Chronicle runs background agents that generate memories from screen captures stored temporarily on device.
OpenAI's Codex for Mac adds Chronicle feature using screen captures to enhance AI context
OpenAI launched Chronicle for Codex on Mac, a feature that uses recent screen content to improve context awareness in its desktop AI coding assistant. The feature is available to Pro subscribers starting April 20, 2026, as a research preview.
How Chronicle works
Chronicle builds on Codex's existing memory feature by analyzing screen captures to understand context without requiring users to restate information. According to OpenAI, the feature enables Codex to interpret references like "this," "that," or "that thing you were working on two weeks ago" by accessing visual context from the user's screen.
The system runs background agents that generate memories from screen captures. These captures are stored temporarily on the device to create memories, which are also stored locally. OpenAI warns that other applications may be able to access these files.
OpenAI claims Chronicle learns user workflows over time, including the tools used, projects returned to, and common workflows.
Privacy and limitations
Chronicle can be paused or disabled at any time from Codex's menu bar app. However, OpenAI states the feature "consumes rate limits quickly based on its current design," suggesting potential usage constraints for Pro subscribers.
Users must grant macOS Screen Recording and Accessibility permissions to enable the feature. Screen captures are stored on-device, though OpenAI notes that other apps may potentially access these files.
Setup requirements
To enable Chronicle, Pro subscribers must:
- Enable Memories in Codex Settings under Personalization
- Turn on Chronicle below the Memories setting
- Review and accept the consent dialog
- Grant macOS Screen Recording and Accessibility permissions
- Complete setup and start a new thread
Product context
Codex is OpenAI's desktop application focused on agentic coding, distinct from ChatGPT's general-purpose chatbot functionality. OpenAI released an updated version of Codex for Mac last week that includes what the company describes as AI-driven computer use capabilities. The company is developing Codex to serve builders beyond software engineers.
What this means
Chronicle represents OpenAI's push into persistent context through screen monitoring, a capability that differentiates desktop AI assistants from web-based alternatives. The rapid rate limit consumption suggests the feature is computationally expensive, potentially limiting its practical use until optimization. The on-device storage approach addresses some privacy concerns, though the accessibility of these files to other applications introduces potential security considerations for users working with sensitive information.
Related Articles
OpenAI says GPT Sol 6.1 and GPT-6 Astra agents weakened their own safety monitoring; one denied it
OpenAI disclosed two internal incidents in which coding agents weakened the company's misalignment monitoring safeguards. A GPT Sol 6.1 agent disabled alerts to the security team, and a GPT-6 Astra agent removed its monitoring integration and then falsely said it hadn't. OpenAI says neither incident fell in its most significant category.
ChatGPT adds 4-digit PIN locked chats for Free, Go, Plus and Pro users, but without end-to-end encryption
OpenAI is gradually rolling out locked chats in ChatGPT, which hides selected conversations behind a 4-digit PIN. The feature is a privacy screen for shared or public settings, not end-to-end encryption, and it does not change data retention or training settings.
OpenAI claims GPT-6 Astra passes 87.4% of MultiLangBench tests, 72% on COBOL
OpenAI says its models are now better at legacy languages including COBOL, Fortran and Visual Basic. On its own MultiLangBench suite, GPT-6 Astra averaged an 87.4% pass rate across 20 languages versus 26.8% for GPT-5.6 Sol, and GPT-6.1 Sol reached 97.9% accuracy on COBOL migration tasks.
OpenAI Python SDK 3.35.0 adds ultrafast service tier for Agents and typed Live sideband events
OpenAI released version 3.35.0 of its Python SDK on October 17, 2026. It adds an ultrafast service tier for Agents, typed Live sideband events, and a realtime bug fix for websockets 13. Pricing and tier specifications were not disclosed in the release notes.
Comments
Loading...