Field guide / primary domain
39
sources in this field
Updated August 31, 2026
Current thesis
The shortest path to orientation.
This synthesis records claims and practices from the cited sources; reported outcomes and product capabilities have not been independently verified.
Codex orchestrates multi-agent work by routing Claude for architectural reasoning and specialized executors for deterministic tasks, reducing wasted agent PRs from ~50% to 0%. Record & Replay converts demonstrated workflows into inspectable skills; self-managing threads create work autonomously. A four-step clarification prompt surfaces hidden gaps, and constraining models to ask exactly one question increases user response likelihood. Codex-Orchestration enables per-role model assignment; Fable 5 planner + Sol executor solved problems both Opus and GPT-5.5 struggled with in 30 minutes at 40% fewer rate-limit hits. Luna Max performs roughly on par with Opus 5 Medium at ~1/6th cost. Beyond scripting, Codex enables durable personal automation: processing voice/note output into structured notes, auto-parsing emails into task managers, and auto-tagging read-later items. Viticci reports automation workflows built 3 months prior remain in daily use unchanged, suggesting well-scoped personal automation has longer shelf life than expected. Reusable /goal templates span software maintenance audits, performance benchmarking, UI polish, and ideation. A Codex user reported receiving ad hoc quota boosts of 10%-50%. The prompt 'Organize all my recent chats into relevant sections' effectively clusters prior chat history into usable sections.
Evidence board
Claims worth carrying forward01Codex prompt tip: 'Organize all my recent chats into relevant sections' is reported to effectively cluster and categorize prior chat history into usable sections.
02A Codex user (@thsottiaux) reported receiving a usage reset plus 10%-50% more usage allowance, suggesting Codex sometimes grants ad hoc quota boosts/resets to active users.
03Codex for Mac can be used to automatically process voice/note app output (e.g., from Monologue) into structured notes, demonstrating durable personal automation beyond one-off scripting.
04A working Codex automation pattern: parse incoming emails and auto-save relevant ones into a task manager, removing manual triage as a recurring workflow.
05Codex automation applied to read-later apps: auto-tagging saved items based on content, reducing manual organization overhead for personal knowledge management.
06Viticci reports that automations built 3 months prior in a Codex guide are still in daily use unchanged, suggesting well-scoped personal automation workflows have longer shelf life than expected in a fast-moving AI tooling landscape.
07Subrouter is a local Go proxy that routes Codex/Claude Code traffic across multiple ChatGPT Pro/Claude Max subscriptions and API keys, using sticky conversation-to-account assignment so cached context stays useful.
08Subrouter's account selection policy scores accounts by most constrained usage window, protects low-headroom accounts for new sessions, spends quota resetting soonest, and prefers subscription OAuth accounts over API-key accounts when tied.
Adjacent fields
Latest evidence
Dan McAteer
A tip circulating among Codex users: running the prompt 'Organize all my recent chats into relevant sections' effectively categorizes chat history, paired with a mention of a usage reset/boost (10-50% more) from another user.
Federico Viticci
Federico Viticci highlights recurring Codex-based automations on macOS he built and still uses months later, including processing voice notes, email-to-task routing, and auto-tagging read-later items, detailed in a MacStories Club guide.
Ассизи
Subrouter is a Go-based local proxy (manaflow-ai) that distributes Codex and Claude Code traffic across multiple ChatGPT Pro/Claude Max subscriptions and API keys, using sticky conversation-to-account assignment, usage-aware selection, multi-tenant isolation, and optional full-transcript recording.
Vox
A reusable reasoning prompt forces Codex or Claude to surface hidden assumptions, missing information, and common mistakes before answering, then ask exactly one clarifying question and wait for a reply—improving response quality by front-loading disambiguation.
Forward Future Brian
A collection of high-effort /goal prompt templates designed to consume leftover Codex quota before a reset, spanning software cleanup, performance auditing, personal life planning, and skill-building.
Aaron Klein 🇺🇸
Author reports that Claude and Codex agents sometimes generate invented hyper-technical terminology and incoherent output, and that adding a specific instruction to AGENTS.md (shown in an attached image, not transcribed in the text) resolved the issue.
Ben Davis
A config change lets Codex's multiple-choice tool function outside plan mode, reportedly improving usability significantly, though the exact config snippet was shared as an image rather than text.
Dan McAteer
A free open-source Codex plugin called 'sol-advisor' implements a four-role orchestration pattern using different GPT-5.6 model tiers to manage cost and capability tradeoffs while claiming to avoid weekly usage limits.
Sarah Chieng
OpenAI's jxnlco describes a Codex pattern using thread managers—threads that communicate with other threads via Slack context, hourly loops, and feedback IDs—where the real benefit is better-scoped context per thread, not just parallel execution.
Alvaro Cintas
A new open-source Codex plugin, Codex-Orchestration, lets users assign different models (e.g., Fable 5, Sol) to specific roles like advisor or executor, with Codex handling routing between them. Author reports 40% fewer limit hits and 2x faster implementation on a task set that stumped other models.