pi-provider-kimi-code 0.6.2 → 0.6.5
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 +7 -5
- package/index.ts +308 -469
- package/package.json +4 -4
- package/src/config.ts +18 -5
- package/src/constants.ts +2 -2
- package/src/device.ts +41 -0
- package/src/models.ts +100 -28
- package/src/oauth.ts +137 -35
- package/src/payload.ts +38 -25
- package/src/project-trust.ts +24 -0
- package/src/settings-ui.ts +204 -0
- package/src/stream.ts +25 -6
- package/src/tools/common.ts +6 -4
- package/src/tools/moonshot.ts +13 -35
- package/src/usage.ts +366 -0
package/README.md
CHANGED
|
@@ -102,16 +102,18 @@ KIMI_API_KEY=sk-... pi
|
|
|
102
102
|
|
|
103
103
|
## Model
|
|
104
104
|
|
|
105
|
-
This provider publishes
|
|
105
|
+
This provider publishes two Pi model IDs:
|
|
106
106
|
|
|
107
107
|
```text
|
|
108
108
|
kimi-coding/kimi-for-coding
|
|
109
|
+
kimi-coding/kimi-for-coding-highspeed
|
|
109
110
|
```
|
|
110
111
|
|
|
111
|
-
Select
|
|
112
|
+
Select standard or highspeed mode inside Pi:
|
|
112
113
|
|
|
113
114
|
```text
|
|
114
115
|
/model kimi-coding/kimi-for-coding
|
|
116
|
+
/model kimi-coding/kimi-for-coding-highspeed
|
|
115
117
|
```
|
|
116
118
|
|
|
117
119
|
Kimi keeps Coding models behind aliases. Rather than hardcoding a stale model list, this extension asks Kimi for the current model info when you log in or refresh. If your account is on a newer rollout (e.g., Kimi K2.7) or internal test, Pi can pick up the latest model name and context size without waiting for a package release.
|
|
@@ -123,7 +125,7 @@ Fallback values:
|
|
|
123
125
|
- Input: text and image
|
|
124
126
|
- Reasoning: enabled
|
|
125
127
|
|
|
126
|
-
The provider maps Pi's reasoning levels to Kimi's `
|
|
128
|
+
The provider maps Pi's reasoning levels to Kimi's top-level `thinking` parameter. It sends `thinking.effort` only when `/models` advertises supported effort values. The mapping refreshes automatically on credential refresh. Opening `/kimi-settings` also re-discovers the latest model metadata.
|
|
127
129
|
|
|
128
130
|
## Optional tools
|
|
129
131
|
|
|
@@ -135,7 +137,7 @@ Inside Pi, run:
|
|
|
135
137
|
/kimi-settings
|
|
136
138
|
```
|
|
137
139
|
|
|
138
|
-
That command shows the current server-side model name (e.g. "K2.7 Code High Speed"), your Kimi
|
|
140
|
+
That command shows the current server-side model name (e.g. "K2.7 Code High Speed"), your Kimi quota and Extra Usage balance, and lets you edit the home or project config. Changes apply to the active session tool set.
|
|
139
141
|
|
|
140
142
|
Configurable settings include protocol mode, upload threshold, and per-tool enable/collapse.
|
|
141
143
|
|
|
@@ -277,7 +279,7 @@ Kimi's cache is **content-based**: it fires automatically when your prompt prefi
|
|
|
277
279
|
- A timestamp, request ID, or randomized header is being interpolated into the prompt.
|
|
278
280
|
- The first ~256 tokens of the prompt differ between turns.
|
|
279
281
|
|
|
280
|
-
`PI_CACHE_RETENTION=none` skips `prompt_cache_key` injection but **does not** disable Kimi's caching (the cache is unconditional). For deterministic measurement, run `scripts/
|
|
282
|
+
`PI_CACHE_RETENTION=none` skips `prompt_cache_key` injection but **does not** disable Kimi's caching (the cache is unconditional). For deterministic measurement, run a focused suite under `scripts/e2e/cache/`, starting with `scripts/e2e/cache/ttl.sh`. See [docs/caching.md](docs/caching.md) for full mechanics.
|
|
281
283
|
|
|
282
284
|
### OpenAI-compatible tools complain about a `developer` role
|
|
283
285
|
|