# Portable Agent Plugins Arrive as CI Becomes the Bottleneck *By Coding Agents Alpha Tracker • August 7, 2026* OpenAI, Vercel, Cursor, and their partners are turning skills and MCP configurations into portable agent plugins. The same period’s GitHub Actions outage is a reminder that elegant agent loops still terminate at infrastructure. ## 🔥 TOP SIGNAL **Agent Plugins makes the harness portable.** OpenAI’s announcement describes an open standard built with AWS, Cursor, GitHub, VS Code, and Vercel that packages Agent Skills and MCP server configurations in a shared format; Romain Huet frames it as the portability step after `AGENTS.md`, Agent Skills, and `.agents` config. [^1][^2] Cursor says it already supports bundling skills and MCP servers across agents. [^3] Build one small skill-plus-MCP bundle and test it in two clients now; interoperability is the hypothesis to validate, not a benefit to assume. ## ⚡ TRY THIS - **Turn parallel threads into a dependency graph.** Create one thread per independently testable task, have it ping back when finished, and preserve each thread’s work rather than collapsing everything into one context. swyx says this creates a workable kanban/waterfall graph in today’s coding agents; Theo’s inbox-style sidebar is the UI version of the same idea—he says it lets him keep 4+ threads active without losing context. [^4][^5] - **Choose the harness before writing the prompt.** LangChain’s current rule is to start with `create_deep_agent` when you want the bundled context machinery: filesystem-backed context, subagents, on-demand skills, and memory. Use the minimal LangChain loop plus middleware when you need fine-grained tool/context control or deterministic summarization and verification; move to LangGraph when the workflow needs a custom graph, durable execution, fault tolerance, or human-in-the-loop steps. [^6] - **Route models by task, then measure the routing.** Cursor’s vendor-reported starting matrix is Grok 4.5 for routine work, GPT-5.6 Sol for planning and codebase comprehension, Opus 5 for execution-heavy tasks, and Fable 5 for debugging and visual implementation. Treat that as a policy to test against your own repository with latency, cost, and acceptance checks—not as an independent benchmark. [^7][^8] - **Run a product-shaped eval instead of another toy benchmark.** swyx proposed giving developers $1,000 in tokens and one weekend to clone an enterprise SaaS his team might otherwise buy for more than $40,000 per year; the prospective customer evaluates the result, the winner gets $10,000, and the code is open-sourced. Borrow the format at smaller scale: pick a workflow your team actually cares about, cap tokens and time, and let the would-be user judge the build. This is a proposed eval format, not a demonstrated result. [^9] ## 📡 WHAT SHIPPED - **Codex Security Review — research preview.** OpenAI says it uses repository context to inspect GitHub pull requests for security issues and place actionable findings directly in the PR; automatic reviews are documented here: [security review setup](https://learn.chatgpt.com/docs/security/security-review). Greg Brockman says Codex can now review every GitHub pull request and leave findings inline. [^10][^11] - **pi 0.84.0.** Armin Ronacher’s agent tool adds fullscreen/alt-screen mode, LaTeX and Mermaid rendering in the transcript, improved Windows support, and `AGENTS.override.md` support. [Full changelog.](https://pi.dev/news/releases/0.84.0) [^12][^13] - **Cloudflare MCP v2.** The rewritten MCP core is stateless and designed to run on web infrastructure such as Workers, with a new feature lifecycle and SDK migration path. Kent C. Dodds says Kody Koala is already updated to the latest spec, calling the change a major simplification. [^14][^15] - **Smol Forge opened its first 100-user alpha.** swyx describes it as a fast, agent-native Git remote; the onboarding instruction is to point an agent at `forge.smol.ai/llms.txt`. It is explicitly rough—transcript functionality is broken—so this is a hands-on watchlist item, not a production recommendation. [^16] - **Omarchy Quattro moved from alpha to beta.** DHH reports a codebase 4× the size of the 3-series, including roughly 40K lines of Quickshell QML, while the overall system is a quarter of its former code, the ISO is 20% smaller, and installs are up to 40% faster. He calls it his most agent-accelerated codebase and credits frontier models with QML writing and Linux diagnosis; those are maintainer-reported figures, not an independent benchmark. [^17][^18][^19] - **bb is picking up an adoption signal.** Ben Tossell calls it “likely my new default.” Sawyer Hood’s project is an open-source, MIT-licensed orchestrator that works with Codex, Claude Code, Cursor, and ACP-compatible agents, while letting users ask the agent to extend the IDE itself. [Repo.](https://github.com/get-bb/bb) [^20][^21] - **CI reality check:** Theo reported GitHub Actions had been down for five hours, with 85% of Actions webhooks not being processed; Armin Ronacher said a pi patch release was blocked because publishing depended on CI. Check provider status before debugging an agent loop that is waiting on PR webhooks or deployment signals. [^22][^23][^24] ## 🎬 GO DEEPER - **Video — [Riley Brown: Guillermo Rauch on Vercel’s internal agent](https://www.youtube.com/watch?v=HQXi4snP36I).** The episode covers Vercel’s internal agent V, described as being used by almost 1,000 people, then gets concrete about building an agent: start with an `instructions.md` identity, add a `tools/` folder, connect a team channel, and keep writes behind existing permissions. [^25] [![The AI Agent Every Company is About to Build | Vercel CEO Guillermo Rauch](https://img.youtube.com/vi/HQXi4snP36I/hqdefault.jpg)](https://youtube.com/watch?v=HQXi4snP36I&t=1650) *The AI Agent Every Company is About to Build | Vercel CEO Guillermo Rauch (27:30)* The follow-on segment is the part to copy: collect thumbs-up/down feedback, aggregate failures overnight, keep a human in the improvement loop, and encode behavior and accuracy as evals. [^25] [![The AI Agent Every Company is About to Build | Vercel CEO Guillermo Rauch](https://img.youtube.com/vi/HQXi4snP36I/hqdefault.jpg)](https://youtube.com/watch?v=HQXi4snP36I&t=2318) *The AI Agent Every Company is About to Build | Vercel CEO Guillermo Rauch (38:38)* - **Repo — [bb](https://github.com/get-bb/bb).** Study its extension model rather than just its UI: the launch post lists agent-built task management, 2D thread navigation, GitHub-webhook code review using Codex computer use, markdown editing, crons, inline previews, and remote access. It is a useful specimen of an orchestrator whose missing features become prompts. [^21] **Editorial take:** The durable alpha edge is moving from clever prompts to portable capabilities, task-routed models, and measured control planes. [^1][^6] --- ### Sources [^1]: [𝕏 post by @OpenAIDevs](https://x.com/OpenAIDevs/status/2085398373511918022) [^2]: [𝕏 post by @romainhuet](https://x.com/romainhuet/status/2085535288957513872) [^3]: [𝕏 post by @cursor_ai](https://x.com/cursor_ai/status/2085464617694777762) [^4]: [𝕏 post by @swyx](https://x.com/swyx/status/2085253030417461661) [^5]: [𝕏 post by @theo](https://x.com/theo/status/2085487362453414037) [^6]: [𝕏 article by @sydneyrunkle](https://x.com/i/article/2085420681387442176) [^7]: [𝕏 post by @cursor_ai](https://x.com/cursor_ai/status/2085390483740676365) [^8]: [𝕏 post by @cursor_ai](https://x.com/cursor_ai/status/2085390485502239171) [^9]: [𝕏 post by @swyx](https://x.com/swyx/status/2085517544795079014) [^10]: [𝕏 post by @OpenAIDevs](https://x.com/OpenAIDevs/status/2085482310636560830) [^11]: [𝕏 post by @gdb](https://x.com/gdb/status/2085496677725860064) [^12]: [𝕏 post by @mitsuhiko](https://x.com/mitsuhiko/status/2085324285178962137) [^13]: [𝕏 post by @mitsuhiko](https://x.com/mitsuhiko/status/2085324287368454253) [^14]: [𝕏 post by @mattzcarey](https://x.com/mattzcarey/status/2085352166017937765) [^15]: [𝕏 post by @kentcdodds](https://x.com/kentcdodds/status/2085383982372540429) [^16]: [𝕏 post by @swyx](https://x.com/swyx/status/2085450774914756631) [^17]: [𝕏 post by @dhh](https://x.com/dhh/status/2085111385495572733) [^18]: [𝕏 post by @dhh](https://x.com/dhh/status/2085307947605758179) [^19]: [𝕏 post by @dhh](https://x.com/dhh/status/2085308432593088685) [^20]: [𝕏 post by @bentossell](https://x.com/bentossell/status/2085300851824787550) [^21]: [𝕏 article by @sawyerhood](https://x.com/i/article/2084798414513115136) [^22]: [𝕏 post by @theo](https://x.com/theo/status/2085481030283386988) [^23]: [𝕏 post by @theo](https://x.com/theo/status/2085488398438162696) [^24]: [𝕏 post by @mitsuhiko](https://x.com/mitsuhiko/status/2085457206887977232) [^25]: [The AI Agent Every Company is About to Build | Vercel CEO Guillermo Rauch](https://www.youtube.com/watch?v=HQXi4snP36I)