Anyone running OpenCode against local models for real work?
Not benchmarks, not demos — daily driving. Qwen3 Coder on a 4090 versus the API bill: where local actually holds up, where it quietly fails, and the routing setup that makes it viable.
@moss.dev
Full-stack generalist. My AGENTS.md is longer than some codebases I've worked in.
Reputation is earned per category — these are narrow, evidence-backed badges, not one karma number.
Not benchmarks, not demos — daily driving. Qwen3 Coder on a 4090 versus the API bill: where local actually holds up, where it quietly fails, and the routing setup that makes it viable.
Mine grew to 600 lines and the agent ignores most of it. Time to talk about what earns a slot in the most expensive real estate in your repo.
Everyone arguing skeleton quality is missing that briefs beat both. My scaffold prompt is 300 lines of AGENTS.md at this point — stack pins,…
on Codex vs Cursor for greenfield scaffolding · 3h ago
Sample-twice-and-diff is delightfully stupid and I implemented it in an hour as an OpenCode plugin (plugin API held still long enough!). Ear…
on Anyone running OpenCode against local models for real work? · 1w ago
For the record, the before/after from my logs:
on OpenCode retry loop hammers rate limits when the provider returns 529 · 1w ago
The 'to make its own code compile' failure mode generalizes way beyond iOS — I've seen tsconfig strictness lowered for the same reason. The…
on Best AGENTS.md for SwiftUI production apps · 2w ago
Leaning no for now: the warning helps discovery, but the underlying position-priority effect is unchanged and the other agents warn about no…
on Oversized AGENTS.md gets silently truncated · 2w ago
Agents love 'fixing' installs by regenerating the lockfile, and every regeneration is a 3,000-line diff nobody reviews. Pin the package manager, pin exact versions, and make CI reject any lockfile the humans didn't ask for.
Our AGENTS.md grew to 1,900 lines of accumulated team wisdom. Somewhere past a few thousand tokens, agents stop seeing the bottom half — rules in the top sections get followed, rules in the bottom sections might as well not exist. No warning from any tool.
Same shape: poetry install --sync (or uv sync --frozen) as the blessed command, plus the same CI check on poetry.lock vs pyproject.toml. The…
on A lockfile policy that stops agent churn · 2w ago