# Jev goes fully open, Claude Code token-saving workshop, Harness Tax paper drops Date: 2026-09-21 Language: en Canonical: https://code-echo.pages.dev/2026/09/21/ai-twitter-hots-en --- # AI Twitter/X Highlights Digest · 2026-09-21 (Mon) ## Key Takeaways 1. **Jev** drops the waitlist for everyone; open **Kev** decision models and **DocJev** land the same day. 2. **Claude Code**’s official “spend fewer tokens” workshop goes viral in JP; **Devin** keeps free SWE-2 in Cloud plus Fusion chatter. 3. **Cline Desktop** hits ~6% of all tasks one week in; **Codex** $200 angst meets Cloudflare’s hot **security-audit-skill**. 4. A UC Berkeley-led look at **Harness Tax**: swap shells, similar success rates, up to ~5× cost—minimal **Pi** wins on spend. --- ## 1. Jev: no waitlist, open to everyone **Summary**: @typesafeai announced **Jev is available to everyone—no waitlist**. The same day, @omarsar0 shipped an auto-updating Awesome Jev Collection and an interactive **Jev Primer & Playground**, with curation powered by Jev itself. **Why it matters**: Classifiers move from invite-only novelty to a default agent decision primitive—the conversation shifts from access to integration. - Open access: https://x.com/typesafeai/status/2101786156572823624 - Awesome Collection: https://x.com/omarsar0/status/2101696753749655863 - Primer & Playground: https://x.com/omarsar0/status/2101775584661573692 ![Jev open](/images/twitter-hots/2026-09-21/01-jev-open.jpg) --- ## 2. Kev: open-source Jev-like decision models you can train **Summary**: @jaredpalmer updated **Kev-0.6B / 4B / 8B**—open (Apache 2.0) Jev-style decision models on Qwen3. Out-of-domain: Kev-8B ~79.6% vs Jev 85.7%. Kev-4B runs five questions in ~300 ms bf16 on a 32 GB Mac; TypeSafe System One SDK works with a `base_url` swap. **Why it matters**: The day Jev opens the gate, a train-it-yourself open alternative appears for the same decision-head niche. - Link: https://x.com/jaredpalmer/status/2101715352258232539 ![Kev models](/images/twitter-hots/2026-09-21/02-kev.jpg) --- ## 3. DocJev: document classify & split with Jev **Summary**: LlamaIndex’s @jerryjliu0 launched open-source **DocJev**: feed a document plus natural-language category rules; Jev classifies or finds sub-document boundaries. Claims ~6× faster than gpt-5.6-luna at similar accuracy. OCR via liteparse (fast) or LlamaParse (complex layouts). **Why it matters**: Decision models land in the ingestion path—not just chat routing. - Link: https://x.com/jerryjliu0/status/2101738281046294552 ![DocJev](/images/twitter-hots/2026-09-21/03-docjev.jpg) --- ## 4. Claude Code: official token-saving workshop + JP skills **Summary**: @ClaudeCode_love amplified Anthropic’s ~26-minute Claude Code token workshop: `/clear` between tasks, route Haiku/Sonnet/Opus/Fable by job, disable idle MCP with `/mcp`, farm heavy research to sub-agents, and enable telemetry for cache hit rate and per-person cost. @minorun365 also open-sourced a Skill for building talk decks with Claude Code. **Why it matters**: Official guidance on not burning context—often more useful than another model SKU. - Workshop: https://x.com/ClaudeCode_love/status/2101538091668484599 - Deck Skill: https://x.com/minorun365/status/2101672853443957238 ![Claude Code tokens](/images/twitter-hots/2026-09-21/04-claude-code.jpg) --- ## 5. Devin: free SWE-2 in Cloud + Fusion + macOS teardown **Summary**: @dabit3 noted **SWE-2** remains free in Devin Cloud Agents / CLI / Desktop through **Oct 8** on paid plans, and argued multi-model **Fusion** harnesses are the future. @b_shulha tried Fusion (Fable + SWE-2) and called it a rare auto-mode with real WOW. @LinearUncle highlighted Cognition engineer Jon’s *Bringing macOS to Devin*—EC2 Mac, Virtualization.framework, accessibility-tree CLI, and computer use. **Why it matters**: Free-flagship window, multi-model orchestration, and cloud Mac engineering details all in one feed day. - Free SWE-2: https://x.com/dabit3/status/2101684714318778414 - Fusion future: https://x.com/dabit3/status/2101714747733254512 - Fusion hands-on: https://x.com/b_shulha/status/2101653743007629778 - macOS teardown: https://x.com/LinearUncle/status/2101582102588502096 ![Devin SWE-2](/images/twitter-hots/2026-09-21/05-devin.jpg) --- ## 6. Cline Desktop: ~6% of tasks after one week **Summary**: @cline said **Cline Desktop** already runs over **6%** of all tasks one week after launch, with about half from brand-new users; **Kimi K3** remains free in-app for a limited time. **Why it matters**: Desktop traction measured in real task share, not another changelog bullet. - Link: https://x.com/cline/status/2101754721677840702 ![Cline Desktop](/images/twitter-hots/2026-09-21/06-cline-desktop.jpg) --- ## 7. Codex: subscription angst + Cloudflare security-audit-skill trending **Summary**: @haydendevs floated canceling a **$200** Codex sub for open models—“can’t justify paying for something I only use a few times a month.” Separately @huhuhang pointed surplus Codex quota at GitHub Trending #1 **cloudflare/security-audit-skill**. The repo was created **2026-06**—today’s spike is recirculation, not a brand-new open-source drop. **Why it matters**: Same product line, two fights at once: is it worth it, and where should spare quota go. - Subscription take: https://x.com/haydendevs/status/2101718661945823336 - security-audit-skill: https://x.com/huhuhang/status/2101502764484624813 ![Codex & audit skill](/images/twitter-hots/2026-09-21/07-codex.jpg) --- ## 8. Harness Tax: similar success, up to ~5× cost **Summary**: @iwashi86 summarized UC Berkeley-led work across **Claude Code, Codex CLI, and minimal OSS Pi** with seven models: swapping harnesses moves success rates only ~±2–5% on average, while token cost for the same model can differ by up to ~**5×**—dubbed **Harness Tax**. Tiny-tool Pi was most cost-efficient; Claude Code’s first-turn context can exceed Pi’s by 10×+. Vendor-native harnesses were not always best (e.g. some Sonnet runs better on Codex CLI; some GPT runs better/cheaper on Pi). **Why it matters**: Half the bill is the shell—not just the model leaderboard. - Link: https://x.com/iwashi86/status/2101808623643197789 ![Harness Tax](/images/twitter-hots/2026-09-21/08-harness-tax.jpg) --- ## 9. OpenCode V2 and Muse Spark 1.3 **Summary**: @Fei2411 said **OpenCode V2** feels much better than v1, with **Gemini 3.8 Flash** and **Step 5** looking stronger in that harness than in Antigravity / DeepSeek Harness / Codex. @alexandr_wang nudged people to try **Muse Spark 1.3** in **OpenCode** or Muse Code. **Why it matters**: Open harnesses and model-lab coding stacks keep cross-endorsing—pick both the model and the shell. - OpenCode V2: https://x.com/Fei2411/status/2101699077096567214 - Muse Spark 1.3: https://x.com/alexandr_wang/status/2101567604204908900 ![OpenCode & Muse](/images/twitter-hots/2026-09-21/09-opencode-muse.jpg) --- ## 10. Pi: Vienna Sunday meditation on agentic SE friction **Summary**: From Vienna, @pidotdev shared a Sunday conversation where @mitsuhiko talks through shared frustrations of agentic software engineering—and reminds listeners they’re not alone. A nice counterpoint to today’s Harness Tax thread: the lightest shell is still talking honestly about how hard this feels. **Why it matters**: Not a release note—harness authors naming why coding with agents still feels awkward. - Link: https://x.com/pidotdev/status/2101672600237646305 ![Pi Vienna](/images/twitter-hots/2026-09-21/10-pi-vienna.jpg)