@datalayer/agent-runtimes 1.3.15 → 1.3.17

package/lib/specs/models.js

@@ -28,6 +28,14 @@ export const AIModels = {
     BEDROCK_US_ANTHROPIC_CLAUDE_SONNET_4_5_20250929_V1_0: 'bedrock:us.anthropic.claude-sonnet-4-5-20250929-v1:0',
     BEDROCK_US_ANTHROPIC_CLAUDE_SONNET_4_6: 'bedrock:us.anthropic.claude-sonnet-4-6',
     BEDROCK_US_ANTHROPIC_CLAUDE_SONNET_4_20250514_V1_0: 'bedrock:us.anthropic.claude-sonnet-4-20250514-v1:0',
+    CLOUDFLARE_GTW_TYPESAFE_JEV: 'cloudflare:gtw/typesafe/jev',
+    CLOUDFLARE_WRK_GOOGLE_GEMMA_4_26B_A4B_IT: 'cloudflare:wrk/google/gemma-4-26b-a4b-it',
+    CLOUDFLARE_WRK_ZAI_ORG_GLM_5_2: 'cloudflare:wrk/zai-org/glm-5.2',
+    CLOUDFLARE_WRK_OPENAI_GPT_OSS_120B: 'cloudflare:wrk/openai/gpt-oss-120b',
+    CLOUDFLARE_WRK_MOONSHOTAI_KIMI_K2_6: 'cloudflare:wrk/moonshotai/kimi-k2.6',
+    CLOUDFLARE_WRK_META_LLAMA_3_3_70B_INSTRUCT_FP8_FAST: 'cloudflare:wrk/meta/llama-3.3-70b-instruct-fp8-fast',
+    CLOUDFLARE_WRK_QWEN_QWEN3_8_27B: 'cloudflare:wrk/qwen/qwen3.8-27b',
+    CLOUDFLARE_WRK_TYPESAFE_JEV: 'cloudflare:wrk/typesafe/jev',
     OLLAMA_GEMMA3_4B: 'ollama:gemma3:4b',
     OLLAMA_LLAMA3_1_8B: 'ollama:llama3.1:8b',
     OLLAMA_QWEN2_5_CODER_7B: 'ollama:qwen2.5-coder:7b',
@@ -47,6 +55,7 @@ export const ALIBABA_QWEN_MAX_0_0_1 = {
     name: 'Alibaba Qwen-Max',
     description: 'Qwen-Max via Alibaba Cloud Model Studio - highest capability Qwen model',
     provider: 'alibaba',
+    providerUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
     default: false,
     available: true,
     requiredEnvVars: [],
@@ -59,6 +68,7 @@ export const ALIBABA_QWEN3_32B_0_0_1 = {
     name: 'Alibaba Qwen3-32B',
     description: 'Qwen3-32B via Alibaba Cloud Model Studio - open-weight 32B model with tool calling',
     provider: 'alibaba',
+    providerUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
     default: false,
     available: false,
     requiredEnvVars: [],
@@ -71,6 +81,7 @@ export const ALIBABA_QWEN3_6_FLASH_0_0_1 = {
     name: 'Alibaba Qwen3.6-Flash',
     description: 'Qwen3.6-Flash via Alibaba Cloud Model Studio - fast and low cost',
     provider: 'alibaba',
+    providerUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
     default: false,
     available: false,
     requiredEnvVars: [],
@@ -83,6 +94,7 @@ export const ALIBABA_QWEN3_6_PLUS_0_0_1 = {
     name: 'Alibaba Qwen3.6-Plus',
     description: 'Qwen3.6-Plus via Alibaba Cloud Model Studio - balanced performance and cost',
     provider: 'alibaba',
+    providerUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
     default: false,
     available: false,
     requiredEnvVars: [],
@@ -95,6 +107,7 @@ export const ALIBABA_QWEN3_7_PLUS_0_0_1 = {
     name: 'Alibaba Qwen3.7-Plus',
     description: 'Qwen3.7-Plus via Alibaba Cloud Model Studio - balanced flagship Qwen3 model',
     provider: 'alibaba',
+    providerUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
     default: false,
     available: false,
     requiredEnvVars: [],
@@ -107,6 +120,7 @@ export const ANTHROPIC_CLAUDE_3_5_HAIKU_20241022_0_0_1 = {
     name: 'Anthropic Claude Haiku 3.5',
     description: 'Claude Haiku 3.5 by Anthropic - fast and efficient',
     provider: 'anthropic',
+    providerUrl: 'https://www.anthropic.com/claude/haiku',
     default: false,
     available: false,
     requiredEnvVars: ['ANTHROPIC_API_KEY'],
@@ -118,6 +132,7 @@ export const ANTHROPIC_CLAUDE_OPUS_4_20250514_0_0_1 = {
     name: 'Anthropic Claude Opus 4',
     description: 'Claude Opus 4 by Anthropic - highest capability model',
     provider: 'anthropic',
+    providerUrl: 'https://www.anthropic.com/claude/opus',
     default: false,
     available: false,
     requiredEnvVars: ['ANTHROPIC_API_KEY'],
@@ -129,6 +144,7 @@ export const ANTHROPIC_CLAUDE_SONNET_4_5_20250514_0_0_1 = {
     name: 'Anthropic Claude Sonnet 4.5',
     description: 'Claude Sonnet 4.5 by Anthropic - balanced performance and speed',
     provider: 'anthropic',
+    providerUrl: 'https://www.anthropic.com/claude/sonnet',
     default: false,
     available: false,
     requiredEnvVars: ['ANTHROPIC_API_KEY'],
@@ -140,6 +156,7 @@ export const ANTHROPIC_CLAUDE_SONNET_4_20250514_0_0_1 = {
     name: 'Anthropic Claude Sonnet 4',
     description: 'Claude Sonnet 4 by Anthropic - strong reasoning and coding',
     provider: 'anthropic',
+    providerUrl: 'https://www.anthropic.com/claude/sonnet',
     default: false,
     available: false,
     requiredEnvVars: ['ANTHROPIC_API_KEY'],
@@ -151,6 +168,7 @@ export const AZURE_OPENAI_GPT_4_1_MINI_0_0_1 = {
     name: 'Azure OpenAI GPT-4.1 Mini',
     description: 'GPT-4.1 Mini via Azure OpenAI - compact version',
     provider: 'azure-openai',
+    providerUrl: 'https://learn.microsoft.com/en-us/azure/ai-foundry/foundry-models/concepts/models-sold-directly-by-azure',
     default: false,
     available: false,
     requiredEnvVars: ['AZURE_OPENAI_API_KEY', 'AZURE_OPENAI_ENDPOINT'],
@@ -162,6 +180,7 @@ export const AZURE_OPENAI_GPT_4_1_NANO_0_0_1 = {
     name: 'Azure OpenAI GPT-4.1 Nano',
     description: 'GPT-4.1 Nano via Azure OpenAI - smallest and fastest',
     provider: 'azure-openai',
+    providerUrl: 'https://learn.microsoft.com/en-us/azure/ai-foundry/foundry-models/concepts/models-sold-directly-by-azure',
     default: false,
     available: false,
     requiredEnvVars: ['AZURE_OPENAI_API_KEY', 'AZURE_OPENAI_ENDPOINT'],
@@ -173,6 +192,7 @@ export const AZURE_OPENAI_GPT_4_1_0_0_1 = {
     name: 'Azure OpenAI GPT-4.1',
     description: 'GPT-4.1 via Azure OpenAI - strong general purpose',
     provider: 'azure-openai',
+    providerUrl: 'https://learn.microsoft.com/en-us/azure/ai-foundry/foundry-models/concepts/models-sold-directly-by-azure',
     default: false,
     available: false,
     requiredEnvVars: ['AZURE_OPENAI_API_KEY', 'AZURE_OPENAI_ENDPOINT'],
@@ -184,6 +204,7 @@ export const AZURE_OPENAI_GPT_4O_MINI_0_0_1 = {
     name: 'Azure OpenAI GPT-4o Mini',
     description: 'GPT-4o Mini via Azure OpenAI - compact enterprise deployment',
     provider: 'azure-openai',
+    providerUrl: 'https://learn.microsoft.com/en-us/azure/ai-foundry/foundry-models/concepts/models-sold-directly-by-azure',
     default: false,
     available: false,
     requiredEnvVars: ['AZURE_OPENAI_API_KEY', 'AZURE_OPENAI_ENDPOINT'],
@@ -195,6 +216,7 @@ export const AZURE_OPENAI_GPT_4O_0_0_1 = {
     name: 'Azure OpenAI GPT-4o',
     description: 'GPT-4o via Azure OpenAI - enterprise deployment',
     provider: 'azure-openai',
+    providerUrl: 'https://learn.microsoft.com/en-us/azure/ai-foundry/foundry-models/concepts/models-sold-directly-by-azure',
     default: false,
     available: false,
     requiredEnvVars: ['AZURE_OPENAI_API_KEY', 'AZURE_OPENAI_ENDPOINT'],
@@ -206,6 +228,7 @@ export const BEDROCK_US_ANTHROPIC_CLAUDE_FABLE_5_0_0_1 = {
     name: 'Bedrock Claude Fable 5',
     description: 'Claude Fable 5 via AWS Bedrock',
     provider: 'bedrock',
+    providerUrl: 'https://aws.amazon.com/bedrock/anthropic/',
     default: false,
     available: false,
     requiredEnvVars: [
@@ -221,6 +244,7 @@ export const BEDROCK_US_ANTHROPIC_CLAUDE_OPUS_4_6_V1_0_0_1 = {
     name: 'Bedrock Claude Opus 4.6',
     description: 'Claude Opus 4.6 via AWS Bedrock',
     provider: 'bedrock',
+    providerUrl: 'https://aws.amazon.com/bedrock/anthropic/',
     default: false,
     available: false,
     requiredEnvVars: [
@@ -236,6 +260,7 @@ export const BEDROCK_US_ANTHROPIC_CLAUDE_OPUS_4_8_0_0_1 = {
     name: 'Bedrock Claude Opus 4.8',
     description: 'Claude Opus 4.8 via AWS Bedrock',
     provider: 'bedrock',
+    providerUrl: 'https://aws.amazon.com/bedrock/anthropic/',
     default: false,
     available: false,
     requiredEnvVars: [
@@ -251,6 +276,7 @@ export const BEDROCK_US_ANTHROPIC_CLAUDE_OPUS_4_20250514_V1_0_0_0_1 = {
     name: 'Bedrock Claude Opus 4',
     description: 'Claude Opus 4 via AWS Bedrock - highest capability',
     provider: 'bedrock',
+    providerUrl: 'https://aws.amazon.com/bedrock/anthropic/',
     default: false,
     available: false,
     requiredEnvVars: [
@@ -266,6 +292,7 @@ export const BEDROCK_US_ANTHROPIC_CLAUDE_OPUS_5_0_0_1 = {
     name: 'Bedrock Claude Opus 5',
     description: 'Claude Opus 5 via AWS Bedrock - the current frontier model',
     provider: 'bedrock',
+    providerUrl: 'https://aws.amazon.com/bedrock/anthropic/',
     default: false,
     available: false,
     requiredEnvVars: [
@@ -281,6 +308,7 @@ export const BEDROCK_US_ANTHROPIC_CLAUDE_SONNET_4_5_20250929_V1_0_0_0_1 = {
     name: 'Bedrock Claude Sonnet 4.5',
     description: 'Claude Sonnet 4.5 via AWS Bedrock - balanced performance',
     provider: 'bedrock',
+    providerUrl: 'https://aws.amazon.com/bedrock/anthropic/',
     default: false,
     available: false,
     requiredEnvVars: [
@@ -296,6 +324,7 @@ export const BEDROCK_US_ANTHROPIC_CLAUDE_SONNET_4_6_0_0_1 = {
     name: 'Bedrock Claude Sonnet 4.6',
     description: 'Claude Sonnet 4.6 via AWS Bedrock - balanced performance',
     provider: 'bedrock',
+    providerUrl: 'https://aws.amazon.com/bedrock/anthropic/',
     default: true,
     available: true,
     requiredEnvVars: [
@@ -304,6 +333,8 @@ export const BEDROCK_US_ANTHROPIC_CLAUDE_SONNET_4_6_0_0_1 = {
         'AWS_DEFAULT_REGION',
     ],
     tokensLimit: 64000,
+    capabilities: ['chat', 'tools', 'judge'],
+    contextWindow: 200000,
 };
 export const BEDROCK_US_ANTHROPIC_CLAUDE_SONNET_4_20250514_V1_0_0_0_1 = {
     id: 'bedrock:us.anthropic.claude-sonnet-4-20250514-v1:0',
@@ -311,6 +342,7 @@ export const BEDROCK_US_ANTHROPIC_CLAUDE_SONNET_4_20250514_V1_0_0_0_1 = {
     name: 'Bedrock Claude Sonnet 4',
     description: 'Claude Sonnet 4 via AWS Bedrock - strong reasoning',
     provider: 'bedrock',
+    providerUrl: 'https://aws.amazon.com/bedrock/anthropic/',
     default: false,
     available: false,
     requiredEnvVars: [
@@ -320,12 +352,151 @@ export const BEDROCK_US_ANTHROPIC_CLAUDE_SONNET_4_20250514_V1_0_0_0_1 = {
     ],
     tokensLimit: 64000,
 };
+export const CLOUDFLARE_GTW_TYPESAFE_JEV_0_0_1 = {
+    id: 'cloudflare:gtw/typesafe/jev',
+    version: '0.0.1',
+    name: 'Jev (Cloudflare AI Gateway)',
+    description: "Typesafe's typed-judgment model through the account's AI Gateway - noul, choice and score questions answered as calibrated probabilities; 32k context; zero data retention; the gateway keeps the logs and bills from its credits",
+    provider: 'cloudflare',
+    providerUrl: 'https://docs.typesafe.ai/models',
+    default: false,
+    available: true,
+    requiredEnvVars: [],
+    capabilities: ['judgments'],
+    billing: 'credits',
+    route: 'ai-gateway',
+    contextWindow: 32000,
+    zeroDataRetention: true,
+    requestLogging: 'gateway',
+    pricing: { inputUsdPerMillion: 0.042, outputUsdPerMillion: 0.0 },
+};
+export const CLOUDFLARE_WRK_GOOGLE_GEMMA_4_26B_A4B_IT_0_0_1 = {
+    id: 'cloudflare:wrk/google/gemma-4-26b-a4b-it',
+    version: '0.0.1',
+    name: 'Cloudflare Gemma 4 26B',
+    description: 'Google Gemma 4 26B (a4b, instruction-tuned) on Cloudflare Workers AI - 256k context, tool calling, standard billing',
+    provider: 'cloudflare',
+    providerUrl: 'https://developers.cloudflare.com/workers-ai/models/gemma-4-26b-a4b-it/',
+    default: false,
+    available: false,
+    requiredEnvVars: [],
+    tokensLimit: 16384,
+    capabilities: ['chat', 'tools', 'codemode'],
+    billing: 'standard',
+    route: 'workers-ai',
+    contextWindow: 256000,
+    aliases: ['cloudflare:google/gemma-4-26b-a4b-it'],
+};
+export const CLOUDFLARE_WRK_ZAI_ORG_GLM_5_2_0_0_1 = {
+    id: 'cloudflare:wrk/zai-org/glm-5.2',
+    version: '0.0.1',
+    name: 'Cloudflare GLM-5.2',
+    description: 'Z.ai GLM-5.2 on Cloudflare Workers AI - 262k context, tool calling; a frontier model billed from AI Gateway credits',
+    provider: 'cloudflare',
+    providerUrl: 'https://developers.cloudflare.com/workers-ai/models/glm-5.2/',
+    default: false,
+    available: false,
+    requiredEnvVars: [],
+    tokensLimit: 32768,
+    capabilities: ['chat', 'tools', 'codemode'],
+    billing: 'credits',
+    route: 'workers-ai',
+    contextWindow: 262144,
+    aliases: ['cloudflare:zai-org/glm-5.2'],
+};
+export const CLOUDFLARE_WRK_OPENAI_GPT_OSS_120B_0_0_1 = {
+    id: 'cloudflare:wrk/openai/gpt-oss-120b',
+    version: '0.0.1',
+    name: 'Cloudflare gpt-oss-120b',
+    description: 'OpenAI gpt-oss-120b on Cloudflare Workers AI - 128k context, tool calling, standard billing',
+    provider: 'cloudflare',
+    providerUrl: 'https://developers.cloudflare.com/workers-ai/models/gpt-oss-120b/',
+    default: false,
+    available: false,
+    requiredEnvVars: [],
+    tokensLimit: 32768,
+    capabilities: ['chat', 'tools', 'codemode', 'judge'],
+    billing: 'standard',
+    route: 'workers-ai',
+    contextWindow: 128000,
+    aliases: ['cloudflare:openai/gpt-oss-120b'],
+};
+export const CLOUDFLARE_WRK_MOONSHOTAI_KIMI_K2_6_0_0_1 = {
+    id: 'cloudflare:wrk/moonshotai/kimi-k2.6',
+    version: '0.0.1',
+    name: 'Cloudflare Kimi K2.6',
+    description: 'Moonshot Kimi K2.6 on Cloudflare Workers AI - 262k context, tool calling; a frontier model billed from AI Gateway credits',
+    provider: 'cloudflare',
+    providerUrl: 'https://developers.cloudflare.com/workers-ai/models/kimi-k2.6/',
+    default: false,
+    available: false,
+    requiredEnvVars: [],
+    tokensLimit: 32768,
+    capabilities: ['chat', 'tools', 'codemode'],
+    billing: 'credits',
+    route: 'workers-ai',
+    contextWindow: 262144,
+    aliases: ['cloudflare:moonshotai/kimi-k2.6'],
+};
+export const CLOUDFLARE_WRK_META_LLAMA_3_3_70B_INSTRUCT_FP8_FAST_0_0_1 = {
+    id: 'cloudflare:wrk/meta/llama-3.3-70b-instruct-fp8-fast',
+    version: '0.0.1',
+    name: 'Cloudflare Llama 3.3 70B',
+    description: 'Meta Llama 3.3 70B Instruct (fp8, fast) on Cloudflare Workers AI - 24k context, tool calling, standard billing',
+    provider: 'cloudflare',
+    providerUrl: 'https://developers.cloudflare.com/workers-ai/models/llama-3.3-70b-instruct-fp8-fast/',
+    default: false,
+    available: false,
+    requiredEnvVars: [],
+    tokensLimit: 8192,
+    capabilities: ['chat', 'tools', 'codemode'],
+    billing: 'standard',
+    route: 'workers-ai',
+    contextWindow: 24000,
+    aliases: ['cloudflare:meta/llama-3.3-70b-instruct-fp8-fast'],
+};
+export const CLOUDFLARE_WRK_QWEN_QWEN3_8_27B_0_0_1 = {
+    id: 'cloudflare:wrk/qwen/qwen3.8-27b',
+    version: '0.0.1',
+    name: 'Cloudflare Qwen3.8 27B',
+    description: 'Qwen3.8 27B on Cloudflare Workers AI - 262k context, tool calling, standard billing',
+    provider: 'cloudflare',
+    providerUrl: 'https://developers.cloudflare.com/workers-ai/models/qwen3.8-27b/',
+    default: false,
+    available: false,
+    requiredEnvVars: [],
+    tokensLimit: 16384,
+    capabilities: ['chat', 'tools', 'codemode'],
+    billing: 'standard',
+    route: 'workers-ai',
+    contextWindow: 262144,
+    aliases: ['cloudflare:qwen/qwen3.8-27b'],
+};
+export const CLOUDFLARE_WRK_TYPESAFE_JEV_0_0_1 = {
+    id: 'cloudflare:wrk/typesafe/jev',
+    version: '0.0.1',
+    name: 'Jev (Cloudflare Workers AI)',
+    description: "Typesafe's typed-judgment model at Workers AI's own endpoint - noul, choice and score questions answered as calibrated probabilities; 32k context; zero data retention; billed from the account's credits, no gateway in the way",
+    provider: 'cloudflare',
+    providerUrl: 'https://docs.typesafe.ai/models',
+    default: false,
+    available: true,
+    requiredEnvVars: [],
+    capabilities: ['judgments'],
+    billing: 'credits',
+    route: 'workers-ai',
+    contextWindow: 32000,
+    zeroDataRetention: true,
+    requestLogging: 'none',
+    pricing: { inputUsdPerMillion: 0.042, outputUsdPerMillion: 0.0 },
+};
 export const OLLAMA_GEMMA3_4B_0_0_1 = {
     id: 'ollama:gemma3:4b',
     version: '0.0.1',
     name: 'Gemma 3 4B (Ollama)',
     description: 'Gemma 3 4B running locally through Ollama - small and fast, no tool calling',
     provider: 'ollama',
+    providerUrl: 'https://ollama.com/library/gemma3:4b',
     default: false,
     available: false,
     requiredEnvVars: [],
@@ -339,6 +510,7 @@ export const OLLAMA_LLAMA3_1_8B_0_0_1 = {
     name: 'Llama 3.1 8B (Ollama)',
     description: 'Meta Llama 3.1 8B running locally through Ollama - tool calling, no data leaves the machine',
     provider: 'ollama',
+    providerUrl: 'https://ollama.com/library/llama3.1:8b',
     default: false,
     available: false,
     requiredEnvVars: [],
@@ -352,6 +524,7 @@ export const OLLAMA_QWEN2_5_CODER_7B_0_0_1 = {
     name: 'Qwen2.5 Coder 7B (Ollama)',
     description: 'Qwen2.5 Coder 7B running locally through Ollama - code-focused with tool calling',
     provider: 'ollama',
+    providerUrl: 'https://ollama.com/library/qwen2.5-coder:7b',
     default: false,
     available: false,
     requiredEnvVars: [],
@@ -365,6 +538,7 @@ export const OPENAI_GPT_4_1_MINI_0_0_1 = {
     name: 'OpenAI GPT-4.1 Mini',
     description: 'GPT-4.1 Mini by OpenAI - compact version of GPT-4.1',
     provider: 'openai',
+    providerUrl: 'https://platform.openai.com/docs/models/gpt-4.1-mini',
     default: false,
     available: false,
     requiredEnvVars: ['OPENAI_API_KEY'],
@@ -376,6 +550,7 @@ export const OPENAI_GPT_4_1_NANO_0_0_1 = {
     name: 'OpenAI GPT-4.1 Nano',
     description: 'GPT-4.1 Nano by OpenAI - smallest and fastest',
     provider: 'openai',
+    providerUrl: 'https://platform.openai.com/docs/models/gpt-4.1-nano',
     default: false,
     available: false,
     requiredEnvVars: ['OPENAI_API_KEY'],
@@ -387,6 +562,7 @@ export const OPENAI_GPT_4_1_0_0_1 = {
     name: 'OpenAI GPT-4.1',
     description: 'GPT-4.1 by OpenAI - strong general purpose model',
     provider: 'openai',
+    providerUrl: 'https://platform.openai.com/docs/models/gpt-4.1',
     default: false,
     available: false,
     requiredEnvVars: ['OPENAI_API_KEY'],
@@ -398,6 +574,7 @@ export const OPENAI_GPT_4O_MINI_0_0_1 = {
     name: 'OpenAI GPT-4o Mini',
     description: 'GPT-4o Mini by OpenAI - compact and cost-effective',
     provider: 'openai',
+    providerUrl: 'https://platform.openai.com/docs/models/gpt-4o-mini',
     default: false,
     available: false,
     requiredEnvVars: ['OPENAI_API_KEY'],
@@ -409,6 +586,7 @@ export const OPENAI_GPT_4O_0_0_1 = {
     name: 'OpenAI GPT-4o',
     description: 'GPT-4o by OpenAI - fast multimodal model',
     provider: 'openai',
+    providerUrl: 'https://platform.openai.com/docs/models/gpt-4o',
     default: false,
     available: false,
     requiredEnvVars: ['OPENAI_API_KEY'],
@@ -420,6 +598,7 @@ export const OPENAI_O3_MINI_0_0_1 = {
     name: 'OpenAI o3 Mini',
     description: 'o3 Mini by OpenAI - reasoning-focused compact model',
     provider: 'openai',
+    providerUrl: 'https://platform.openai.com/docs/models/o3-mini',
     default: false,
     available: false,
     requiredEnvVars: ['OPENAI_API_KEY'],
@@ -451,6 +630,14 @@ export const AI_MODEL_CATALOGUE = {
     'bedrock:us.anthropic.claude-sonnet-4-5-20250929-v1:0': BEDROCK_US_ANTHROPIC_CLAUDE_SONNET_4_5_20250929_V1_0_0_0_1,
     'bedrock:us.anthropic.claude-sonnet-4-6': BEDROCK_US_ANTHROPIC_CLAUDE_SONNET_4_6_0_0_1,
     'bedrock:us.anthropic.claude-sonnet-4-20250514-v1:0': BEDROCK_US_ANTHROPIC_CLAUDE_SONNET_4_20250514_V1_0_0_0_1,
+    'cloudflare:gtw/typesafe/jev': CLOUDFLARE_GTW_TYPESAFE_JEV_0_0_1,
+    'cloudflare:wrk/google/gemma-4-26b-a4b-it': CLOUDFLARE_WRK_GOOGLE_GEMMA_4_26B_A4B_IT_0_0_1,
+    'cloudflare:wrk/zai-org/glm-5.2': CLOUDFLARE_WRK_ZAI_ORG_GLM_5_2_0_0_1,
+    'cloudflare:wrk/openai/gpt-oss-120b': CLOUDFLARE_WRK_OPENAI_GPT_OSS_120B_0_0_1,
+    'cloudflare:wrk/moonshotai/kimi-k2.6': CLOUDFLARE_WRK_MOONSHOTAI_KIMI_K2_6_0_0_1,
+    'cloudflare:wrk/meta/llama-3.3-70b-instruct-fp8-fast': CLOUDFLARE_WRK_META_LLAMA_3_3_70B_INSTRUCT_FP8_FAST_0_0_1,
+    'cloudflare:wrk/qwen/qwen3.8-27b': CLOUDFLARE_WRK_QWEN_QWEN3_8_27B_0_0_1,
+    'cloudflare:wrk/typesafe/jev': CLOUDFLARE_WRK_TYPESAFE_JEV_0_0_1,
     'ollama:gemma3:4b': OLLAMA_GEMMA3_4B_0_0_1,
     'ollama:llama3.1:8b': OLLAMA_LLAMA3_1_8B_0_0_1,
     'ollama:qwen2.5-coder:7b': OLLAMA_QWEN2_5_CODER_7B_0_0_1,
@@ -461,5 +648,33 @@ export const AI_MODEL_CATALOGUE = {
     'openai:gpt-4o': OPENAI_GPT_4O_0_0_1,
     'openai:o3-mini': OPENAI_O3_MINI_0_0_1,
 };
+/**
+ * The ids a model answered to before its id moved, to the id it has now.
+ * Kept apart from the catalogue, which lists each model once.
+ */
+export const AI_MODEL_ALIASES = {
+    'cloudflare:google/gemma-4-26b-a4b-it': 'cloudflare:wrk/google/gemma-4-26b-a4b-it',
+    'cloudflare:zai-org/glm-5.2': 'cloudflare:wrk/zai-org/glm-5.2',
+    'cloudflare:openai/gpt-oss-120b': 'cloudflare:wrk/openai/gpt-oss-120b',
+    'cloudflare:moonshotai/kimi-k2.6': 'cloudflare:wrk/moonshotai/kimi-k2.6',
+    'cloudflare:meta/llama-3.3-70b-instruct-fp8-fast': 'cloudflare:wrk/meta/llama-3.3-70b-instruct-fp8-fast',
+    'cloudflare:qwen/qwen3.8-27b': 'cloudflare:wrk/qwen/qwen3.8-27b',
+};
+/** A model by its id, or by an id it had before; undefined when the catalogue has neither. */
+export function getModel(modelId) {
+    return AI_MODEL_CATALOGUE[AI_MODEL_ALIASES[modelId] ?? modelId];
+}
+/**
+ * A model a chat can run on: not a typed-judgment model (Jev), which
+ * answers typed questions about a state and nothing else. A spec that
+ * states no capability is read as a chat model.
+ */
+export function isChatModel(model) {
+    return !(model.capabilities ?? []).includes('judgments');
+}
+/** The models a person choosing a chat model is offered from. */
+export function listChatModels() {
+    return Object.values(AI_MODEL_CATALOGUE).filter(isChatModel);
+}
 export const DEFAULT_MODEL = AIModels.BEDROCK_US_ANTHROPIC_CLAUDE_SONNET_4_6;
 export const DEFAULT_MODEL_SPEC = BEDROCK_US_ANTHROPIC_CLAUDE_SONNET_4_6_0_0_1;

package/lib/specs/ops.d.ts

@@ -0,0 +1,15 @@
+/**
+ * Op Catalog.
+ *
+ * Orchestrated, supervised workflows: Cogs, and how their work is verified.
+ * Every Op is resolved: its Guards, its Gates and its Track are in it.
+ *
+ * This file is AUTO-GENERATED from YAML specifications.
+ * DO NOT EDIT MANUALLY - run 'make specs' to regenerate.
+ */
+import type { OpSpec } from '../types/agentspecs';
+export declare const OP_SALES_PIPELINE_BOARD_REPORT_0_0_1: OpSpec;
+export declare const OP_CATALOGUE: Record<string, OpSpec>;
+/** An Op, by `id` or `id:version`, or undefined. */
+export declare function getOp(ref: string): OpSpec | undefined;
+export declare function listOps(): OpSpec[];