@nanobpm/nano-coder 0.29.1 → 0.30.0

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.
Files changed (2) hide show
  1. package/README.md +28 -2
  2. package/package.json +5 -5
package/README.md CHANGED
@@ -368,7 +368,7 @@ Create `~/.config/nano-coder/config.toml` (every field is optional). Directories
368
368
  ```toml
369
369
  model = "anthropic/claude-sonnet-4-5" # provider/model
370
370
  default_provider = "mock" # used when the model has no known provider prefix
371
- temperature = 0.7
371
+ temperature = 0.7 # or "default" to send none (see Temperature below)
372
372
  max_tokens = 4096
373
373
  max_iterations = 0 # LLM calls per user input (0 = unbounded)
374
374
  system_prompt = "You are a helpful assistant with access to tools."
@@ -481,6 +481,30 @@ or your workspace domain override `base_url`, e.g.
481
481
  as thinking models like K3 require. Set `extra_body = { reasoning_effort = "low" }` to
482
482
  make K3 think less.
483
483
 
484
+ ### Temperature
485
+
486
+ `temperature` is a number, or `"default"` to send none so the model uses its own default.
487
+ It can be set for all models, for one provider, or for one model; the most specific
488
+ setting wins:
489
+
490
+ ```toml
491
+ temperature = 0.7 # all models
492
+
493
+ [providers.groq]
494
+ temperature = "default" # every groq model uses its own default
495
+
496
+ [providers.anthropic.models."claude-sonnet-4-5"]
497
+ temperature = 0.3 # this model only
498
+ ```
499
+
500
+ Some models accept no temperature, and for them the model default is the only option:
501
+ GitHub Copilot's reasoning models (GPT-5 and later, Grok, …) and any provider with
502
+ `drop_params = ["temperature"]` (such as the `kimi` preset). A number set for such a
503
+ provider or model is ignored with a warning at startup and when you switch to it; the
504
+ top-level `temperature` just doesn't apply to them. Anthropic accepts 0 to 1, so a higher
505
+ value is sent as 1, with a warning. `/context` shows the temperature in use and where it
506
+ comes from, and `/settings` edits it for the current model, its provider, or all models.
507
+
484
508
  Other per-provider fields: `replay_reasoning`, `max_tokens_param` (`max_tokens`, or `max_completion_tokens`
485
509
  which is the `openai` default), `retry_initial_backoff_ms`, `retry_max_backoff_ms` and
486
510
  `retryable_statuses`.
@@ -634,7 +658,9 @@ sets the window). The window a server has *loaded* is preferred over the model's
634
658
 
635
659
  Compaction asks the current model to summarize older messages, keeping the recent tail
636
660
  (up to 20k tokens, never starting at a tool result). Auto-compaction runs before a model
637
- call when usage passes the threshold. It won't run again until the context has grown by
661
+ call when usage passes the threshold, or earlier if the prompt would leave less than the
662
+ reserved output room (`MIN_OUTPUT_RESERVE`, plus an estimation margin) within the window.
663
+ It won't run again until the context has grown by
638
664
  another 10% of the window, so a context that can't shrink isn't summarized on every call.
639
665
  If summarizing fails, the older messages are dropped with a note. The session log records
640
666
  the new conversation, so `--resume` continues from it.
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@nanobpm/nano-coder",
3
- "version": "0.29.1",
3
+ "version": "0.30.0",
4
4
  "description": "A 6MB coding agent. Run a fleet on your laptop.",
5
5
  "license": "Apache-2.0",
6
6
  "repository": {
@@ -31,9 +31,9 @@
31
31
  "node": ">=18"
32
32
  },
33
33
  "optionalDependencies": {
34
- "@nanobpm/nano-coder-darwin-arm64": "0.29.1",
35
- "@nanobpm/nano-coder-darwin-x64": "0.29.1",
36
- "@nanobpm/nano-coder-linux-arm64": "0.29.1",
37
- "@nanobpm/nano-coder-linux-x64": "0.29.1"
34
+ "@nanobpm/nano-coder-darwin-arm64": "0.30.0",
35
+ "@nanobpm/nano-coder-darwin-x64": "0.30.0",
36
+ "@nanobpm/nano-coder-linux-arm64": "0.30.0",
37
+ "@nanobpm/nano-coder-linux-x64": "0.30.0"
38
38
  }
39
39
  }