Workflows, questions, and show-your-work threads from the community.
Sorry English. I pipe llama 3.3 70b through Groq for agent loops because latency look nice on demos. Yesterday: 900 tok/s for first 4 calls, then hard 429 for 40 minutes while I still had half a Type…
had Cascade refactor a Next.js route handler in Windsurf last night. it opened like 14 related files, rewrote half my zod schema, then hit undo on the one edit that actually fixed the 500. the diff l…
Spent Tuesday wiring Continue to a local qwen2.5 coder via ollama. Autocomplete is fine. The chat panel kept suggesting pip install packages that don't exist on PyPI anymore. Had to roll back two com…
Was debugging a flaky CI job in Cursor and the filesystem MCP happily read my playwright report/trace.zip . 2.1GB. Into the context. Session went sideways, bill jumped $14 before I noticed the spinne…
Edited a 22 min founder interview in Descript last night on my M2. Text edit was fine — cut the ums, moved a paragraph, done. Overdub though. I typed a one line fix and it inserted this weird throat…
running mistral small:24b on my 3090 via ollama. asked for { "status": "ok" | "error", "latency ms": number } with their JSON mode on. got perfectly valid JSON. every time. fields were state , took m…
Was cleaning a stale Next.js branch on my M2 Air. Warp's agent bar suggested chaining rm rf node modules && git push force because "deps looked corrupted". They weren't. The pnpm lock was fine. I alm…
pulled deepseek r1:32b on my M2 Pro last night "just to see". Activity Monitor hit 18GB before the first decent answer showed up, fans sounded like a leaf blower aimed at my face. Q4 was usable for a…
Tried Firefly for a client pitch deck Tuesday. Prompted for muted navy UI mock with soft shadows and got three near identical hero shots with the same fake button labels. The commercially safe angle…
Tried Nightfall’s MCP gateway in front of Cursor this week after one of our agents quietly called a shell MCP that dumped .env into a PR description. It caught a filesystem read outside the allowlist…
Debugging a flaky Remix loader tonight and Phind kept pointing me at a 2022 SO answer about getLoaderData that… doesn’t exist in Remix v2. Worse, it confidently linked a GitHub path that 404s. I almo…
Asked Cascade to "clean up the helm values for staging". It did. Also rewrote three chart templates, renamed a secret key that GitHub Actions still references, and left a cheerful TODO in values prod…
Tried Nightfall MCP Gateway on Tuesday night because Claude Code kept calling gh against the wrong org during a payout hotfix. It did catch the bad call. Then it also blocked the right call for about…
Switched our nightly prompt eval from OpenAI to llama 3.3 70b on Groq last week. First run finished in 41 seconds. Same suite used to take 11 minutes. I spent twenty minutes poking the cache layer be…
Tried Zapier Central this week to draft a Slack→HubSpot deal sync. It looked fine in the preview, then on first live run it invented a "closed won" path that never existed in our pipeline and wrote 1…
Turned on Copilot in Outlook this week because our IT team wouldn’t stop nagging. I asked it to “summarize this thread and propose a next step.” It summarized fine. Then it booked a 30 min 1:1 with s…
Asked Roo Code to tidy auth in my FastAPI app. It renamed /login and /refresh , then left the tests pointing at the old paths. VS Code on Mac. No crash. Diff looked fine until pytest screamed 404 on…
VS Code + Continue on my Electron app. I keep a .continuerules with "never touch package.json scripts without asking". Reload window this morning and boom — rules file still there, but the chat acted…
Tried swapping Claude Sonnet out of our nightly eval harness for mistral small latest on the API. 120 cases, mostly JSON schema + a few tool call traces. Finished in 18 min. Claude was $11 last week…
Ran a 180 case prompt eval against llama 3.3 70b on Groq this morning. First 90 finished in under 4 minutes which felt illegal. Then the free tier hard stopped me mid batch. Error was just rate limit…