agent-ultramode 0.1.0 → 0.1.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +6 -2
- package/package.json +1 -1
package/README.md
CHANGED
|
@@ -1,5 +1,9 @@
|
|
|
1
1
|
# ultra
|
|
2
2
|
|
|
3
|
+
[](https://www.npmjs.com/package/agent-ultramode)
|
|
4
|
+
[](https://github.com/maverick-tr/agent-ultramode/releases)
|
|
5
|
+
[](./LICENSE)
|
|
6
|
+
|
|
3
7
|
Took the [LLM-as-a-Verifier](https://github.com/llm-as-a-verifier/llm-as-a-verifier) paper (Kwok et al., 2026) and turned it into a simple `/ultra` command for coding agents.
|
|
4
8
|
|
|
5
9
|
Instead of shipping the agent's first attempt, `ultra` runs your task **N times in isolated git worktrees**, in parallel, then uses the **same model** as a verifier to pick the best result. When it is confident (or when most attempts succeeded) it applies the winning diff to your working tree; when the attempts diverge and it is not sure, it hands you the top candidates. No cross-model dependency, no extra services, no first-attempt lottery.
|
|
@@ -8,9 +12,9 @@ And unlike most "add a verifier" posts, I benchmarked it before believing it. Th
|
|
|
8
12
|
|
|
9
13
|
## Demo
|
|
10
14
|
|
|
11
|
-
<video src="https://github.com/
|
|
15
|
+
<video src="https://github.com/user-attachments/assets/134487dc-1d60-434b-a2dc-97cfd5942959" controls muted playsinline width="100%"></video>
|
|
12
16
|
|
|
13
|
-
|
|
17
|
+
[▶ Watch the demo](https://github.com/maverick-tr/agent-ultramode/blob/main/ultramode.mp4) (55s, opencode + DeepSeek V4 Flash)
|
|
14
18
|
|
|
15
19
|
## The receipts
|
|
16
20
|
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "agent-ultramode",
|
|
3
|
-
"version": "0.1.
|
|
3
|
+
"version": "0.1.1",
|
|
4
4
|
"description": "Best-of-N for coding agents with a same-model verifier: run your task N times in isolated git worktrees and apply the verified winner. A /ultra command for opencode.",
|
|
5
5
|
"module": "ultra.ts",
|
|
6
6
|
"main": "ultra.ts",
|