Codex
Named by 246 kept items in this workspace.
187 items name it under frameworks and tools, 39 under classification.technologies, 14 under models and 6 under classification.key entities.
The items that name it 246
OpenAI coding agent; usage limits doubled for Pro 100/200/500 and Business Premium.
Coding agent that connects to kanzo through MCP and sees the same tasks as the user.
Coding agent that Ruflo wraps with agent swarms, vector memory and an MCP server.
Harness used to run GPT-6.1 Sol agents in Mindustry.
OpenAI's coding agent. All internal Codex traffic is monitored, regardless of model, and paging is currently disabled for it.
Coding agent running on the user's Mac, which Chums connects to using the existing setup.
Coding agent that Orca runs side by side with others, one worktree each
OpenAI's coding agent. All internal Codex traffic is monitored regardless of model, and paging is currently disabled for it.
AI coding agent named as one of the agents covered by ADLC metrics.
Coding agent that gets the Expo skills from its official plugin marketplace, with automatic updates.
Alternative coding tool for implementing planned chunks. Shares a usage pool with Codex-related chats as described in the page.
Coding agent to switch to once a game project passes the proof-of-concept phase.
Coding agent that can work inside a real analysis environment through MCP bridges
Coding agent that can run the heditor skill to edit videos.
Coding agent run inside BibCode.
Coding agent that read a planted AGENTS.md instruction and followed it on the first try more than 90% of the time this week; it had a rogue MCP server added to its config.
OpenAI coding agent whose usage allowance differs by model; Terra gives twice the usage of Sol.
Coding agent tested in a controlled study of iterative refactoring and over-compliance
OpenAI's coding tool.
OpenAI product where GPT-6.1 Terra is now available.
Coding agent that works with ollama launch.
OpenAI coding product where GPT-6.1 Terra is available to Plus, Pro, Enterprise and Edu users
OpenAI's coding agent. Auto-review is built into it, and internal Codex traffic was used for a deployment simulation.
OpenAI subscription agentic coding harness with usage-based plans.
Agent client that HiRoute can route model requests for.
Coding agent that can read the GenOffice skill file to build a .pptx without opening a window.
Coding agent used by SemiAnalysis to work out the Neuron Explorer profiling setup.
Coding agent that Impeccable works with.
OpenAI's coding agent; internal Codex agents were monitored for misalignment, and METR found a way agents could in principle bypass Codex sandboxing and Auto-Review.
Coding agent that Weco is said to work with.
Agent for which developers wrote skill loaders.
Coding agent that can connect to the Omnara MCP server with codex mcp add omnara --url https://app.omnara.com/mcp.
Coding agent whose engineer sessions were analysed to design the plugins; also the baseline agent in the regression-hunt comparison.
Named as another harness startups might be built on.
Named as a harness on which startups might be built.
OpenAI's coding agent, used internally. Auto-Review is required for all internal use.
Coding agent the ads team used to build and scale the ads system, with guardrails for safe code shipping and data infrastructure plugins to speed insight gathering and improvements.
Coding agent that Anatomy works with.
OpenAI coding agent used for computer-use tasks and for reviewing Opus output.
Coding agent that reads .agents/skills.
OpenAI coding agent harness used for the GPT-6 runs.
GPT-5.5
AI coding tool whose sessions Beacon saves and shares.
Coding agent whose sessions WhatThePort links to the dev servers they started. A reviewer notes the Codex resolver keeps the newest transcript per working directory.
Baseline system compared against OpenVLHarness at maximum reasoning effort.
Coding agent harness used with GPT 5.5.
Coding-agent CLI that Coral can run as part of a mixed-vendor team.
Coding agent named as usable with decomp.wtf through its MCP tools.
Coding agent that can use the DeepSeek API; its server-side tool calls hit the Search and Fetch APIs and are currently free of charge.
Coding agent that Sphinx works with.
OpenAI coding agent; now reuses context across subagents, caches it longer and routes requests for more cache hits.
Coding tool covered in docs/CLI.md for use with PS5LM.
Coding agent that Jev supports
OpenAI's CLI harness (version 0.162.1); default 272k token window, compression triggered at 90% filled
Coding agent that, on the same Mac, fetches the local 127.0.0.1 link to get the marked screenshot image plus app/window info
Used for later repairs to the BeamSpeak source
Product that receives voice updates from GPT-Live-1.1
Coding agent that works with magpie
OpenAI product for which GPT-Live-1.1 is the default voice model for Plus and higher tier users
Coding agent that Graphicle connects to; past sessions are captured as memory
Coding agent and harness giving agents tools to act, create plans, set objectives, spawn subagents and compact context; used to analyze trajectories and fix bottlenecks in the research environment
Coding agent client that hexvero registers with
Existing coding agent supported by Kylon workspace
AI coding assistant in which the user adds the PDF TO MP3 MCP server via Settings → Plugins → Add MCP server
OpenAI's AI coding assistant that REA hooks up to reverse-engineering tools.
Coding agent used with GPT-6 Astra Extra High to build web-edit.
Coding agent the author offloaded thinking tasks to; noted as good at dashboards showing where quotas go
Commercial coding agent newly integrated into the benchmark
Coding agent that Atticus can watch through its own log (e.g. 'tell me when Codex is done')
Coding agent mentioned as running with approvals off
Coding agent that jevmem also works with
OpenAI coding agent; new Windows sandbox on Microsoft Execution Containers (MXC); Composer predictions suggest next message (beta, Pro)
Coding agent in which Eval Skills are used
AI coding agent; bigarrow installs a skill into ~/.agents/skills (with agents/openai.yaml)
Coding agent in which GPT-6 Astra investigated the browser agent, refactored code and ran experiments; Hidalgo set a /goal overnight and reviewed results in the morning.
MCP client for which the MCP guide shows configuration
Coding agent that Meta's Muse Code competes against
A local coding agent that alook can connect to, giving it an identity/inbox so others can talk to it directly.
OpenAI's coding agent, which LegalOn integrated into its development process and expanded across the organization.
AI coding agent that runs agent skills with the permissions of its user
Coding agent credited as the author of several of the site's tool reviews.
A coding agent harness from OpenAI, cited as an example of enterprises' dependency on frontier labs for agent harnesses.
A coding agent platform that Concord integrates with, allowing its agents to message other agents across tools.
Generated summaries interpreting what each of the four SAE latents represents.
OpenAI's coding agent product; contains the Auto-review protocol and is one of the surfaces where older (September) versions of GPT-6 Sol and Luna remain in use.
OpenAI coding product; its models are not changing in this release.
Used by the author to figure out how to run Parseable and feed it OpenTelemetry traces from Datasette.
Coding agent that agent-sprite-forge works in, with a signed-in CLI fallback for generation
OpenAI coding agent product whose adoption surged partly due to goodwill from OpenAI's generous usage resets.
OpenAI product; eligible text output will get an invisible watermark in the EU.
An AI coding agent positioned as an 'AI operating system' that non-coders can run; built around a place to work, a place to run it, controls, and tools.
OpenAI's coding agent product; subject to a global usage reset
A coding agent CLI that Offrun runs locally, signed in as the user, alongside other coding agents.
An agent harness in which models are used, mentioned as one of the harnesses the new RL training approach supports
OpenAI coding agent; default reasoning level can be adjusted, can ask for clarification while working with GPT-6 Astra, and tasks can be steered; supports Ultrafast
Coding agent tool used to run GPT-6 Astra for the World of Warcraft task.
AI coding tool that Switch configures.
Coding agent harness, discussed as structurally similar to Claude Code and Pi.
A coding agent that can be run alongside code in Rhun, used to draft commit messages and whose sessions can be followed as it works, using an existing subscription
An agent harness in which models are trained with RL on task sets, as used by practitioners
An agent harness in which models can be trained/run, as mentioned in the release
OpenAI's coding agent product, now incorporating Computer Use and App Shots so it can build and then test software itself.
Agent that can be connected to Magnitude with one click.
Coding agent that can pull context from the user's meetings over LokalBot MCP or CLI.
OpenAI coding agent product; gains cloud environments that keep running with laptop closed, refreshed CLI with worktrees and /agents, and Security Cloud
Supported AI coding agent that reads skills from .agents/skills and MCP config from .codex/config.toml.
An AI coding assistant into which the invisible-playwright-mcp plugin can be installed via its plugin marketplace.
OpenAI's coding agent; now available on Linux and on phone, fully in the cloud (Codex Cloud/Codex Remote), with an open-sourced harness; used live to rewrite a backend in Rust and build an iPhone app shown in a simulator.
OpenAI coding product where GPT-6.1 Sol is available, with Ultrafast mode coming
OpenAI's agentic coding product; used as the internal traffic source for a deployment simulation that labels trajectories for misaligned behavior to forecast risk changes from deploying GPT-6.1 Sol.
Tasks that dots start or manage there count toward usage limits
Coding agent; in Squadra it acts as builder or reviewer, but a Codex window called first does not coordinate.
OpenAI's coding agent, younger than Claude Code, paired with new models to become reliable enough for daily use
Coding agent (alongside Claude Code) used to run many of the same video-making skills and pipelines.
A coding agent that the jg skill installer can detect and install the Jevgrep skill into.
Coding agent CLI (Codex CLI) hired to sit at a desk; requires native command-hook support and reports only tool names (so shell commands show as plain typing) and root-session token counts, with cost and API-call count unavailable.
A coding agent (used with GPT-5.6 Sol and GPT-6 Astra) evaluated in the paper for synthesizing programs that generalize across generalized TAMP problem instances
Coding agent that can be connected to Whiteboard from the welcome screen.
OpenAI's coding agent, managed by magpie via config.toml edits and exposed as a provider through a ChatGPT sign-in
OpenAI's coding agent tool; the author now uses GPT-6 Sol as its default model.
Can be used to review code, apply prompt caching improvements, and measure results.
OpenAI coding agent where GPT-6 Sol and Luna are available
OpenAI coding agent/tool used internally by researchers and engineers; its spending growth is used by OpenAI as a metric of AI-driven research acceleration.
被引用为「强大」端的高级用户 coding agent 工具，代表低电池、高上限的取舍
A coding agent client Foremerge integrates with; its MCP registration is user-level and serves every repository once set up ('foremerge setup codex'), resolved from the directory Codex is started in.
Coding agent used as a comparison point; Jive claims to be better in speed and token efficiency compared to it, even on tasks not requiring Jev calls.
Used to create V7 Go workflows through MCP.
a coding agent harness that Sno Station integrates with; used alongside Claude Code in the project's own Duo review workflow
Agent tool supported by the installer and compatible with the plain SKILL.md skill files.
Coding agent client supported out of the box via jev-codex; speaks the Responses API (POST /v1/responses)
coding agent that abide hooks into; requires accepting four abide entries via /hooks, edits through apply_patch which abide reads and judges
One of the agents the geo-sleuth skill can be installed into and run from.
Coding agent supported via plugin marketplace; ask is downgraded to a warning because Codex has no ask yet.
A detected agent harness that also receives a small SkillRanker usage skill during install; native Codex integration needs its own support record.
Coding agent client that can install and connect to jev-review via `npx plugins add ... --target codex` or manual setup in ~/.codex/config.toml.
- [AINews] AEF-1 standard emerges for Third Party Evaluators, as Xai, OpenAI, and Anthropic all cosign
OpenAI coding agent app; gained native support for Arch Linux
AI agent that integrates with EvoOntology via a plugin, exposing $build-ontology, $evolve-ontology, and $explore-ontology commands, preparing questions from user needs, project history, and grounded exploration.
Coding agent supported by headroom wrap and MCP install
Agent that can install the skill (--codex)
OpenAI's coding agent product; internal deployment surface for Astra-based coding agents, subject to blocking alignment evaluations and restricted-access safeguards
coding agent that funes can be added to for memory/recall, and that can write to and read from a funes memory
Agent runtime target for choirboy-prompt; behaves like Claude Code, with a SessionStart hook that sends a short status only and relies on the load-context skill for the lore.
A coding agent/tool whose sandbox lets a command write to /tmp or $TMPDIR, so install roots must be kept out of those directories.
Agent that reads the Agent Skills format and can use the skill.
Coding agent used extensively in developing Halo; training images ship repo-specific agent skills discoverable by it.
Agent scaffold used to generate agentic SFT trajectory data.
Coding agent that can be pointed at the SKILL.md endpoint to register and trade directly, with no UI or clone needed.
Coding agent used to run GPT 5.5; its default mode is called default mode
Local agent engine option for BYOA; used as an agent's brain with fail-closed filesystem, command-network, and subprocess-credential boundaries by default.
Existing agent harness (coding agent) that HarnessRouter turns into a pluggable agent backend; usable as a harness_id in the Responses-compatible API.
One of the AI coding agents antislop installs into and works with, via its own plugin marketplace/install commands.
Coding agent (alternative to Claude Code) that the skills can also run under; install.sh can symlink skills into ~/.codex/skills/ with --codex.
r0crawl_skills 所面向的 AI 代理平台；技能目录被复制到其技能目录 ($env:USERPROFILE\.codex\skills) 以供调用。
OpenAI's coding agent product; continues running the July-release versions of GPT-5.6 Sol and GPT-5.6 Luna rather than the newly launched August update
Coding agent run as a harness
Supported agent backend for LongHorizon-Harness (Codex CLI).
Coding agent that integrates with the toolkit via a transparent local proxy for seamless vision support.
Coding agent that ripwire's one-line installer detects and activates skills for.
Coding agent used to run the automated research loop ("/goal" loop) that built the Solidus compiler backend; described as the star contributor to the project
OpenAI coding agent product used as Florian's primary coding tool and as a comparison point for rating Kimi K3's code quality.
Coding agent interface (shown in a figure) that lets users select GPT-5.6's reasoning effort settings.
Coding agent CLI tool that can be connected to 9router-go, also usable as a subscription-tier provider in the routing chain.
- OpenAI Codex with GPT-5.6 Sol: competitive, zero cheating, one unique Django fix | Blog | Endor Labs
OpenAI's agentic coding tool, a CLI-first agent that reads the local codebase, plans edits and applies them through file-change events
A harness used to manage the agentic substrate—where agents live and operate—coupled with mechanisms to store and retrieve memory
Coding agent product cited as an example of a harness, part of the stabilized coding-agent loop across Claude Code, Codex, OpenCode, Cursor-style agents
coding agent that reads the shared AGENTS.md instruction file Graft writes, and has a user-level config (~/.codex/) that graft init can touch
Coding agent mentioned as a supported host for AURI Agents
Coding agent app with an official plugin marketplace; Superpowers can be installed via the Plugins sidebar or the /plugins command.
External ACP-compatible coding agent (OpenAI Codex CLI) that can optionally be integrated for the Coding stage.
Fast CLI coding agent backend option for worker_fast tasks.
支持 Agent Skills 标准的 Agent 工具，可直接安装该仓库的 Skill
AI coding agent tool with measured per-call token counts; supports file-level retry tracking and codeburn context.
Coding agent whose skills and MCP servers can be reused in Rikugan.
AI coding agent harness that ECC tunes.
Agent named as able to read the SKILL.md skills.
Agent host; Ponytail installed via codex plugin marketplace, with two lifecycle hooks to trust in /hooks
Agent client that connects to KYB via MCP or CLI
Agent supported by REA setup
ChatGPT/Codex OAuth sign-in session optionally reused for voice (experimental); login via codex login
Coding agent/plugin marketplace that can add the mattpocock-skills plugin, updating itself at startup.
Coding agent that Overmind integrates with via MCP configuration in .codex/config.toml
AI coding agent evaluated on the Kotlin Benchmark, used together with the GPT 5.5 xhigh model
A coding agent referenced as an example of agents that accumulate steering instructions (skills, AGENTS.md, prompts) over time; shortens skill descriptions to fit context when too many skills are packaged into a project.
Coding agent tool (OpenAI) the course assumes the learner may use; reported to go from 'unreliable' to 'reliable' in a well-harnessed repository with the same model.
AI coding agent harness that can be run as a seat within an OpenRig rig; OpenRig writes its config.toml, trust hashes, activity relay hooks and launches it with -s workspace-write.
Coding agent host (OpenAI) that can install the paper2agent skill in its personal skill location and invoke it via $paper2agent.
A harness product from OpenAI that runs AI coding agents; enterprises buy it rather than build their own, and it decides which model answers, what the model reads, prompt cache use and which subagents run.
A coding agent evaluated in the paper for acting autonomously on DevOps tasks
In every workspace you can read 372
AI
Security
All technologies in this workspace.