harn-vm 0.10.168

Async bytecode virtual machine for the Harn programming language
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104

# Native decision APIs read 2026-10-15. Harn's text/JSON adapter keeps
# probed 255-choice and 2..10-score bounds unless a route needs less.
# Context ceilings come from the route index, not a chat model's weights.
# https://openrouter.ai/api/v1/models?output_modalities=decisions
# https://ai-gateway.vercel.sh/v1/models

[[provider.openrouter]]
model_match = ["openrouter/respan/span-02", "openrouter/nace-ai/drex-v1.5", "openrouter/nace-ai/drex-v2.0"]
extends = true
decision_protocol = "openrouter_decisions"
decision_question_kinds = ["boolean", "choice", "score"]
decision_limits = { max_choice_options = 255, score_levels_min = 2, score_levels_max = 10, state_window_tokens = 131072, request_window_tokens = 131072 }

[[provider.openrouter]]
model_match = ["openrouter/liquid/d1", "openrouter/cloudflare/clef", "openrouter/cloudflare/clef-omni", "openrouter/inception/mercury-decide"]
extends = true
decision_protocol = "openrouter_decisions"
decision_question_kinds = ["boolean", "choice", "score"]
decision_limits = { max_choice_options = 255, score_levels_min = 2, score_levels_max = 10, state_window_tokens = 65536, request_window_tokens = 65536 }

[[provider.openrouter]]
model_match = ["openrouter/liquid/d1-ultra", "openrouter/microsoft/microsoft-decision-1"]
extends = true
decision_protocol = "openrouter_decisions"
decision_question_kinds = ["boolean", "choice", "score"]
decision_limits = { max_choice_options = 255, score_levels_min = 2, score_levels_max = 10, state_window_tokens = 32768, request_window_tokens = 32768 }

[[provider.openrouter]]
model_match = "openrouter/scaledown/classify-1"
extends = true
decision_protocol = "openrouter_decisions"
decision_question_kinds = ["boolean", "choice", "score"]
decision_limits = { max_choice_options = 255, score_levels_min = 2, score_levels_max = 10, state_window_tokens = 1000000, request_window_tokens = 1000000 }

[[provider.openrouter]]
model_match = "openrouter/cloudflare/clef-flash"
extends = true
decision_protocol = "openrouter_decisions"
decision_question_kinds = ["boolean", "choice", "score"]
decision_limits = { max_choice_options = 255, score_levels_min = 2, score_levels_max = 10, state_window_tokens = 24576, request_window_tokens = 24576 }

[[provider.openrouter]]
model_match = ["openrouter/perplexity/pplx-decider-v1.1-27b", "openrouter/perplexity/pplx-decider-latest"]
extends = true
decision_protocol = "openrouter_decisions"
decision_question_kinds = ["boolean", "choice", "score"]
decision_limits = { max_choice_options = 255, score_levels_min = 2, score_levels_max = 10, state_window_tokens = 262144, request_window_tokens = 262144 }

[[provider.openrouter]]
model_match = "openrouter/openai/gpt-6-luna-decisions"
extends = true
decision_protocol = "openrouter_decisions"
decision_question_kinds = ["boolean", "choice", "score"]
decision_limits = { max_questions = 200, max_choice_options = 255, score_levels_min = 2, score_levels_max = 10, state_window_tokens = 1050000, request_window_tokens = 1050000 }

[[provider.openrouter]]
model_match = "openrouter/togethercomputer/tev1-4b-experimental"
extends = true
decision_protocol = "openrouter_decisions"
decision_question_kinds = ["boolean", "choice", "score"]
# Live rejection: "Tev1 needs between 2 and 20 options per question".
decision_limits = { max_choice_options = 20, score_levels_min = 2, score_levels_max = 10, state_window_tokens = 32768, request_window_tokens = 32768 }

[[provider.openrouter]]
model_match = ["openrouter/upstage/solar-decide", "openrouter/upstage/solar-decide-flash"]
extends = true
decision_protocol = "openrouter_decisions"
decision_question_kinds = ["boolean", "choice", "score"]
# Live 422 reports 26 single-token labels on both Solar routes.
decision_limits = { max_choice_options = 26, score_levels_min = 2, score_levels_max = 10, state_window_tokens = 524288, request_window_tokens = 524288 }

[[provider.openrouter]]
model_match = "openrouter/jaredpalmer/kev-4b"
extends = true
decision_protocol = "openrouter_decisions"
decision_question_kinds = ["boolean", "choice", "score"]
decision_limits = { max_choice_options = 255, score_levels_min = 2, score_levels_max = 10, state_window_tokens = 8192, request_window_tokens = 8192 }

[[provider.vercel_ai_gateway]]
model_match = "convaiinnovations/laya"
extends = true
decision_protocol = "vercel_evaluate"
decision_upstream_provider = "boundless"
decision_question_kinds = ["boolean", "choice", "score"]
# 64 short candidates returned a distribution; 128/255 returned HTTP 413.
# This is a tested adapter bound within the route's small context, not a
# published vendor ceiling. The required ZDR filter still blocks this account.
decision_limits = { max_choice_options = 64, score_levels_min = 2, score_levels_max = 10, state_window_tokens = 8192, request_window_tokens = 8192 }

[[provider.vercel_ai_gateway]]
model_match = "liquid/d1-ultra"
extends = true
decision_protocol = "vercel_evaluate"
decision_upstream_provider = "liquid"
decision_question_kinds = ["boolean", "choice", "score"]
decision_limits = { max_choice_options = 255, score_levels_min = 2, score_levels_max = 10, state_window_tokens = 32768, request_window_tokens = 32768 }

[[provider.vercel_ai_gateway]]
model_match = "perplexity/pplx-decider-v1.1-27b"
extends = true
decision_protocol = "vercel_evaluate"
decision_upstream_provider = "perplexity"
decision_question_kinds = ["boolean", "choice", "score"]
decision_limits = { max_choice_options = 255, score_levels_min = 2, score_levels_max = 10, state_window_tokens = 262144, request_window_tokens = 262144 }