@bacnh85/pi-subagent 0.11.0 → 0.12.2
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/CHANGELOG.md +9 -0
- package/README.md +8 -8
- package/agents/general-purpose.md +2 -2
- package/agents/planner.md +1 -2
- package/agents/reviewer.md +1 -2
- package/agents/scout.md +1 -1
- package/agents/tester.md +1 -1
- package/agents/worker.md +2 -2
- package/package.json +9 -9
package/CHANGELOG.md
CHANGED
|
@@ -1,5 +1,14 @@
|
|
|
1
1
|
# Changelog
|
|
2
2
|
|
|
3
|
+
## 0.12.0 (2026-07-24)
|
|
4
|
+
|
|
5
|
+
### Model routing
|
|
6
|
+
|
|
7
|
+
- Chains are now **free-first** to conserve the metered opencode-go budget: paid DeepSeek moves to the **last** position so free nvidia/openrouter fallbacks are tried first. Previously DeepSeek sat at position 2 and burned quota whenever the free GLM primary rate-limited.
|
|
8
|
+
- **Removed two dead fallbacks** found by live-testing every free model on 2026-07-24: `nvidia/moonshotai/kimi-k2.6` (HTTP 404, NVCF not provisioned for the account) and `nvidia/z-ai/glm-5.2` (timeout). These are non-rate-limit failures, so the rate-limit retry loop did not advance past them — a subagent that reached them died instead of falling through to the working `:free` entries behind them.
|
|
9
|
+
- New chains: scout/tester `glm-5-turbo` → `nvidia/gpt-oss-20b` → `deepseek-v4-flash`; worker/general-purpose `glm-5.1` → `nvidia/mistral-small-4-119b-2603` → `nemotron-3-super:free` → `deepseek-v4-flash`; planner/reviewer `glm-5.2` → `nemotron-3-ultra:free` → `deepseek-v4-pro`.
|
|
10
|
+
- Verified-working free additions: `nvidia/openai/gpt-oss-20b` (1.2s), `nvidia/mistralai/mistral-small-4-119b-2603` (2.6s, 119B reasoning), `openrouter/nvidia/nemotron-3-nano-30b-a3b:free` (1.2s).
|
|
11
|
+
|
|
3
12
|
## 0.11.0 (2026-07-23)
|
|
4
13
|
|
|
5
14
|
### Model routing
|
package/README.md
CHANGED
|
@@ -16,14 +16,14 @@ Requires Node.js >= 20.18.
|
|
|
16
16
|
|
|
17
17
|
| Role | Ordered model preferences | Thinking | Tools |
|
|
18
18
|
| --- | --- | --- | --- |
|
|
19
|
-
| `scout` | `zai-coding-cn/glm-5-turbo` → `
|
|
20
|
-
| `tester` | `zai-coding-cn/glm-5-turbo` → `
|
|
21
|
-
| `worker` | `zai-coding-cn/glm-5.1` → `
|
|
22
|
-
| `general-purpose` | `zai-coding-cn/glm-5.1` → `
|
|
23
|
-
| `planner` | `zai-coding-cn/glm-5.2` → `nvidia/
|
|
24
|
-
| `reviewer` | `zai-coding-cn/glm-5.2` → `nvidia/
|
|
19
|
+
| `scout` | `zai-coding-cn/glm-5-turbo` → `nvidia/openai/gpt-oss-20b` → `opencode-go/deepseek-v4-flash` | off | read, grep, find, ls |
|
|
20
|
+
| `tester` | `zai-coding-cn/glm-5-turbo` → `nvidia/openai/gpt-oss-20b` → `opencode-go/deepseek-v4-flash` | off | read, bash, grep, find, ls |
|
|
21
|
+
| `worker` | `zai-coding-cn/glm-5.1` → `nvidia/mistralai/mistral-small-4-119b-2603` → `openrouter/nvidia/nemotron-3-super-120b-a12b:free` → `opencode-go/deepseek-v4-flash` | medium | read, bash, edit, write, grep, find, ls |
|
|
22
|
+
| `general-purpose` | `zai-coding-cn/glm-5.1` → `nvidia/mistralai/mistral-small-4-119b-2603` → `openrouter/nvidia/nemotron-3-super-120b-a12b:free` → `opencode-go/deepseek-v4-flash` | medium | read, bash, edit, write, grep, find, ls |
|
|
23
|
+
| `planner` | `zai-coding-cn/glm-5.2` → `openrouter/nvidia/nemotron-3-ultra-550b-a55b:free` → `opencode-go/deepseek-v4-pro` | high | read, grep, find, ls |
|
|
24
|
+
| `reviewer` | `zai-coding-cn/glm-5.2` → `openrouter/nvidia/nemotron-3-ultra-550b-a55b:free` → `opencode-go/deepseek-v4-pro` | high | read, grep, find, ls |
|
|
25
25
|
|
|
26
|
-
Each role uses the first authenticated preference available through Pi's model registry, then falls back to the authenticated parent model. Chains are
|
|
26
|
+
Each role uses the first authenticated preference available through Pi's model registry, then falls back to the authenticated parent model. Chains are **free-first** to conserve the metered opencode-go budget: **zai-coding-cn** (free GLM, primary) → free **nvidia** NIM and **openrouter** `:free` models → **opencode-go** (paid DeepSeek, last resort — one per role: `deepseek-v4-flash` for fast/strong-coding, `deepseek-v4-pro` for deep reasoning). opencode-go's GLM models cost ~$1.40/$4.40 per M versus zai-coding-cn's free GLM, so GLM stays on zai-coding-cn. Fallback models were live-verified on 2026-07-24; `nvidia/moonshotai/kimi-k2.6` and `nvidia/z-ai/glm-5.2` return 404/timeout on the user's account and were removed — non-rate-limit failures kill the subagent instead of advancing the chain. User/project agent files remain stronger overrides and may set legacy `model`, ordered `models`, and `thinking`.
|
|
27
27
|
|
|
28
28
|
## Agent files
|
|
29
29
|
|
|
@@ -36,8 +36,8 @@ description: Locate relevant files and symbols
|
|
|
36
36
|
tools: read, grep, find, ls
|
|
37
37
|
models:
|
|
38
38
|
- zai-coding-cn/glm-5-turbo
|
|
39
|
-
- opencode-go/deepseek-v4-flash
|
|
40
39
|
- nvidia/openai/gpt-oss-20b
|
|
40
|
+
- opencode-go/deepseek-v4-flash
|
|
41
41
|
---
|
|
42
42
|
|
|
43
43
|
Return concise evidence with file/symbol anchors.
|
|
@@ -4,9 +4,9 @@ description: General-purpose sub-agent for any delegated task. Use when no speci
|
|
|
4
4
|
tools: read, bash, edit, write, grep, find, ls
|
|
5
5
|
models:
|
|
6
6
|
- zai-coding-cn/glm-5.1
|
|
7
|
-
-
|
|
8
|
-
- nvidia/moonshotai/kimi-k2.6
|
|
7
|
+
- nvidia/mistralai/mistral-small-4-119b-2603
|
|
9
8
|
- openrouter/nvidia/nemotron-3-super-120b-a12b:free
|
|
9
|
+
- opencode-go/deepseek-v4-flash
|
|
10
10
|
thinking: medium
|
|
11
11
|
color: yellow
|
|
12
12
|
---
|
package/agents/planner.md
CHANGED
|
@@ -4,9 +4,8 @@ description: Read-only planning and architecture specialist. Use for consequenti
|
|
|
4
4
|
tools: read, grep, find, ls
|
|
5
5
|
models:
|
|
6
6
|
- zai-coding-cn/glm-5.2
|
|
7
|
-
- nvidia/z-ai/glm-5.2
|
|
8
|
-
- opencode-go/deepseek-v4-pro
|
|
9
7
|
- openrouter/nvidia/nemotron-3-ultra-550b-a55b:free
|
|
8
|
+
- opencode-go/deepseek-v4-pro
|
|
10
9
|
thinking: high
|
|
11
10
|
color: blue
|
|
12
11
|
sandbox: read-only
|
package/agents/reviewer.md
CHANGED
|
@@ -4,9 +4,8 @@ description: Code review specialist. Use for correctness, security, regression,
|
|
|
4
4
|
tools: read, grep, find, ls
|
|
5
5
|
models:
|
|
6
6
|
- zai-coding-cn/glm-5.2
|
|
7
|
-
- nvidia/z-ai/glm-5.2
|
|
8
|
-
- opencode-go/deepseek-v4-pro
|
|
9
7
|
- openrouter/nvidia/nemotron-3-ultra-550b-a55b:free
|
|
8
|
+
- opencode-go/deepseek-v4-pro
|
|
10
9
|
thinking: high
|
|
11
10
|
color: purple
|
|
12
11
|
sandbox: read-only
|
package/agents/scout.md
CHANGED
|
@@ -4,8 +4,8 @@ description: Fast codebase recon that returns compressed context for handoff. Us
|
|
|
4
4
|
tools: read, grep, find, ls
|
|
5
5
|
models:
|
|
6
6
|
- zai-coding-cn/glm-5-turbo
|
|
7
|
-
- opencode-go/deepseek-v4-flash
|
|
8
7
|
- nvidia/openai/gpt-oss-20b
|
|
8
|
+
- opencode-go/deepseek-v4-flash
|
|
9
9
|
thinking: off
|
|
10
10
|
color: cyan
|
|
11
11
|
sandbox: read-only
|
package/agents/tester.md
CHANGED
|
@@ -4,8 +4,8 @@ description: Focused verification agent. Use for cheap routine test, typecheck,
|
|
|
4
4
|
tools: read, bash, grep, find, ls
|
|
5
5
|
models:
|
|
6
6
|
- zai-coding-cn/glm-5-turbo
|
|
7
|
-
- opencode-go/deepseek-v4-flash
|
|
8
7
|
- nvidia/openai/gpt-oss-20b
|
|
8
|
+
- opencode-go/deepseek-v4-flash
|
|
9
9
|
thinking: off
|
|
10
10
|
color: orange
|
|
11
11
|
---
|
package/agents/worker.md
CHANGED
|
@@ -3,9 +3,9 @@ name: worker
|
|
|
3
3
|
description: General-purpose coding agent with full tool access. Use only when explicitly requested for isolated implementation.
|
|
4
4
|
models:
|
|
5
5
|
- zai-coding-cn/glm-5.1
|
|
6
|
-
-
|
|
7
|
-
- nvidia/moonshotai/kimi-k2.6
|
|
6
|
+
- nvidia/mistralai/mistral-small-4-119b-2603
|
|
8
7
|
- openrouter/nvidia/nemotron-3-super-120b-a12b:free
|
|
8
|
+
- opencode-go/deepseek-v4-flash
|
|
9
9
|
thinking: medium
|
|
10
10
|
color: green
|
|
11
11
|
---
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@bacnh85/pi-subagent",
|
|
3
|
-
"version": "0.
|
|
3
|
+
"version": "0.12.2",
|
|
4
4
|
"description": "In-process subagents for Pi with isolated SDK sessions, parallel and chained delegation, and inspectable threads.",
|
|
5
5
|
"type": "module",
|
|
6
6
|
"license": "MIT",
|
|
@@ -48,17 +48,17 @@
|
|
|
48
48
|
"check": "npm run typecheck && npm test"
|
|
49
49
|
},
|
|
50
50
|
"peerDependencies": {
|
|
51
|
-
"@earendil-works/pi-coding-agent": ">=0.80.0 <0.
|
|
52
|
-
"@earendil-works/pi-ai": ">=0.80.0 <0.
|
|
53
|
-
"@earendil-works/pi-agent-core": ">=0.80.0 <0.
|
|
54
|
-
"@earendil-works/pi-tui": ">=0.80.0 <0.
|
|
51
|
+
"@earendil-works/pi-coding-agent": ">=0.80.0 <0.83.0",
|
|
52
|
+
"@earendil-works/pi-ai": ">=0.80.0 <0.83.0",
|
|
53
|
+
"@earendil-works/pi-agent-core": ">=0.80.0 <0.83.0",
|
|
54
|
+
"@earendil-works/pi-tui": ">=0.80.0 <0.83.0",
|
|
55
55
|
"typebox": ">=1.3.0 <2.0.0"
|
|
56
56
|
},
|
|
57
57
|
"devDependencies": {
|
|
58
|
-
"@earendil-works/pi-agent-core": "^0.
|
|
59
|
-
"@earendil-works/pi-ai": "^0.
|
|
60
|
-
"@earendil-works/pi-coding-agent": "^0.
|
|
61
|
-
"@earendil-works/pi-tui": "^0.
|
|
58
|
+
"@earendil-works/pi-agent-core": "^0.82.0",
|
|
59
|
+
"@earendil-works/pi-ai": "^0.82.0",
|
|
60
|
+
"@earendil-works/pi-coding-agent": "^0.82.0",
|
|
61
|
+
"@earendil-works/pi-tui": "^0.82.0",
|
|
62
62
|
"@types/mocha": "^10.0.10",
|
|
63
63
|
"@types/node": "^20.19.43",
|
|
64
64
|
"mocha": "^10.8.2",
|