Class: Legion::Extensions::Llm::Provider

Inherits:
Object
  • Object
show all
Includes:
Cache::Helper, Streaming, Logging::Helper
Defined in:
lib/legion/extensions/llm/provider.rb,
lib/legion/extensions/llm/provider/open_ai_compatible.rb

Overview

Base class for LLM providers.

Defined Under Namespace

Modules: OpenAICompatible

Constant Summary collapse

MODEL_DETAIL_CACHE_SCHEMA_VERSION =
2

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Methods included from Streaming

build_on_data_handler, build_stream_callback, build_stream_error_response, error_chunk?, faraday_1?, handle_data, handle_error_chunk, handle_error_event, handle_failed_response, handle_json_error_chunk, handle_parsed_error, handle_sse, handle_stream, json_error_payload?, parse_error_from_json, parse_streaming_error, persist_failed_response_body, persist_failed_response_custom_body?, persist_failed_response_env_body?, process_stream_chunk, raise_partial_streaming_error, raise_streaming_status_error, stream_response

Constructor Details

#initialize(config) ⇒ Provider

Returns a new instance of Provider.



37
38
39
40
41
# File 'lib/legion/extensions/llm/provider.rb', line 37

def initialize(config)
  @config = config.is_a?(Hash) ? HashConfig.new(config) : config
  ensure_configured!
  @connection = Connection.new(self, @config)
end

Instance Attribute Details

#configObject (readonly)

Returns the value of attribute config.



35
36
37
# File 'lib/legion/extensions/llm/provider.rb', line 35

def config
  @config
end

#connectionObject (readonly)

Returns the value of attribute connection.



35
36
37
# File 'lib/legion/extensions/llm/provider.rb', line 35

def connection
  @connection
end

Class Method Details

.assume_models_exist?Boolean

Returns:

  • (Boolean)


527
528
529
# File 'lib/legion/extensions/llm/provider.rb', line 527

def assume_models_exist?
  false
end

.capabilitiesObject



499
500
501
# File 'lib/legion/extensions/llm/provider.rb', line 499

def capabilities
  nil
end

.configuration_optionsObject



507
508
509
# File 'lib/legion/extensions/llm/provider.rb', line 507

def configuration_options
  []
end

.configuration_requirementsObject



503
504
505
# File 'lib/legion/extensions/llm/provider.rb', line 503

def configuration_requirements
  []
end

.configured?(config) ⇒ Boolean

Returns:

  • (Boolean)


535
536
537
# File 'lib/legion/extensions/llm/provider.rb', line 535

def configured?(config)
  configuration_requirements.all? { |req| config.send(req) }
end

.default_tierObject



515
516
517
# File 'lib/legion/extensions/llm/provider.rb', line 515

def default_tier
  :frontier
end

.default_transportObject



511
512
513
# File 'lib/legion/extensions/llm/provider.rb', line 511

def default_transport
  :http
end

.local?Boolean

Returns:

  • (Boolean)


519
520
521
# File 'lib/legion/extensions/llm/provider.rb', line 519

def local?
  false
end

.nameObject



491
492
493
# File 'lib/legion/extensions/llm/provider.rb', line 491

def name
  to_s.split('::').last
end

.remote?Boolean

Returns:

  • (Boolean)


523
524
525
# File 'lib/legion/extensions/llm/provider.rb', line 523

def remote?
  !local?
end

.resolve_model_id(model_id, config: nil) ⇒ Object

rubocop:disable Lint/UnusedMethodArgument



531
532
533
# File 'lib/legion/extensions/llm/provider.rb', line 531

def resolve_model_id(model_id, config: nil) # rubocop:disable Lint/UnusedMethodArgument
  model_id
end

.slugObject



495
496
497
# File 'lib/legion/extensions/llm/provider.rb', line 495

def slug
  name.downcase
end

Instance Method Details

#api_baseObject

Raises:

  • (NotImplementedError)


43
44
45
# File 'lib/legion/extensions/llm/provider.rb', line 43

def api_base
  raise NotImplementedError
end

#assume_models_exist?Boolean

Returns:

  • (Boolean)


259
260
261
# File 'lib/legion/extensions/llm/provider.rb', line 259

def assume_models_exist?
  self.class.assume_models_exist?
end

#cache_control_prefix_tokensObject



243
244
245
246
247
248
249
# File 'lib/legion/extensions/llm/provider.rb', line 243

def cache_control_prefix_tokens
  if config.respond_to?(:cache_control_prefix_tokens) && config.cache_control_prefix_tokens
    config.cache_control_prefix_tokens
  else
    4
  end
end

#cache_enabled?Boolean

Returns:

  • (Boolean)


230
231
232
233
234
235
236
237
238
239
240
241
# File 'lib/legion/extensions/llm/provider.rb', line 230

def cache_enabled?
  explicit = config.llm_cache_enabled if config.respond_to?(:llm_cache_enabled)

  unless explicit.nil?
    log.debug { "[#{slug}] cache_enabled? source=per_provider value=#{explicit}" }
    return explicit == true
  end

  global = global_prompt_caching_enabled?
  log.debug { "[#{slug}] cache_enabled? source=global value=#{global}" }
  global
end

#cache_instance_keyObject



474
475
476
477
478
479
480
481
482
# File 'lib/legion/extensions/llm/provider.rb', line 474

def cache_instance_key
  if cache_local_instance?
    (respond_to?(:instance_id) ? instance_id : :default).to_s
  else
    require 'digest'
    urls = Array(config_base_url).map { |u| strip_scheme(u).downcase.chomp('/') }.sort
    Digest::SHA256.hexdigest(urls.join('|'))[0, 12]
  end
end

#cache_local_instance?Boolean

── Cache helpers with local/shared tier selection ────────────────

Returns:

  • (Boolean)


438
439
440
441
442
443
# File 'lib/legion/extensions/llm/provider.rb', line 438

def cache_local_instance?
  Array(config_base_url).any? do |url|
    host = url.to_s.downcase
    host.include?('localhost') || host.include?('127.0.0.1') || host.include?('::1')
  end
end

#capabilitiesObject



78
79
80
# File 'lib/legion/extensions/llm/provider.rb', line 78

def capabilities
  self.class.capabilities
end

#chat(messages:, model:, tools: [], temperature: nil, params: {}, headers: {}, schema: nil, thinking: nil, tool_prefs: nil) ⇒ Object

rubocop:disable Metrics/ParameterLists



87
88
89
90
# File 'lib/legion/extensions/llm/provider.rb', line 87

def chat(messages:, model:, tools: [], temperature: nil, params: {}, headers: {}, schema: nil, thinking: nil,
         tool_prefs: nil)
  complete(messages, tools:, temperature:, model:, params:, headers:, schema:, thinking:, tool_prefs:)
end

#complete(messages, tools:, temperature:, model:, params: {}, headers: {}, schema: nil, thinking: nil, tool_prefs: nil) ⇒ Object



97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
# File 'lib/legion/extensions/llm/provider.rb', line 97

def complete(messages, tools:, temperature:, model:, params: {}, headers: {}, schema: nil, thinking: nil,
             tool_prefs: nil, &)
  normalized_temperature = maybe_normalize_temperature(temperature, model)
  log_provider_request(
    messages: messages,
    tools: tools,
    temperature: temperature,
    normalized_temperature: normalized_temperature,
    model: model,
    params: params,
    headers: headers,
    schema: schema,
    thinking: thinking,
    tool_prefs: tool_prefs,
    streaming: block_given?
  )

  payload = Utils.deep_merge(
    render_payload(
      messages,
      tools: tools,
      tool_prefs: tool_prefs,
      temperature: normalized_temperature,
      model: model,
      stream: block_given?,
      schema: schema,
      thinking: thinking
    ),
    params
  )

  if block_given?
    stream_response @connection, payload, headers, &
  else
    sync_response @connection, payload, headers
  end
end

#config_base_urlObject



396
397
398
# File 'lib/legion/extensions/llm/provider.rb', line 396

def config_base_url
  respond_to?(:settings) ? settings[:base_url] : nil
end

#configuration_requirementsObject



82
83
84
# File 'lib/legion/extensions/llm/provider.rb', line 82

def configuration_requirements
  self.class.configuration_requirements
end

#configured?Boolean

Returns:

  • (Boolean)


226
227
228
# File 'lib/legion/extensions/llm/provider.rb', line 226

def configured?
  configuration_requirements.all? { |req| @config.send(req) }
end

#count_tokens(messages:, model:, params: {}) ⇒ Object



211
212
213
214
215
216
217
# File 'lib/legion/extensions/llm/provider.rb', line 211

def count_tokens(messages:, model:, params: {})
  _ = [model, params]
  Array(messages).sum do |message|
    content = message.respond_to?(:content) ? message.content : message[:content] || message['content']
    estimate_text_tokens(content)
  end
end

#discover_offerings(live: false, raise_on_unreachable: false, **filters) ⇒ Object



142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
# File 'lib/legion/extensions/llm/provider.rb', line 142

def discover_offerings(live: false, raise_on_unreachable: false, **filters)
  return filter_cached_offerings(Array(@cached_offerings), filters) unless live

  provider_health = health(live:)
  @cached_offerings = Array(list_models(live:, **filters)).filter_map do |model|
    next unless model_matches_filters?(model, filters)
    next unless model_allowed?(model.id)

    offering_from_model(model, health: provider_health)
  end
  @cached_offerings
rescue Faraday::ConnectionFailed, Faraday::TimeoutError => e
  log.warn("[#{slug}] instance=#{provider_instance_id} unreachable: #{e.message}")
  raise if raise_on_unreachable

  []
end

#embed(text:, model:, dimensions: nil, params: {}, headers: {}) ⇒ Object



186
187
188
189
190
191
192
# File 'lib/legion/extensions/llm/provider.rb', line 186

def embed(text:, model:, dimensions: nil, params: {}, headers: {})
  payload = Utils.deep_merge(render_embedding_payload(text, model:, dimensions:), params)
  response = @connection.post(embedding_url(model:), payload) do |req|
    req.headers = headers.merge(req.headers) unless headers.empty?
  end
  parse_embedding_response(response, model:, text:)
end

#endpoint_manifestObject



285
286
287
288
289
290
291
292
293
294
# File 'lib/legion/extensions/llm/provider.rb', line 285

def endpoint_manifest
  endpoint_methods.each_with_object({}) do |(key, method_name), result|
    next unless respond_to?(method_name)

    value = public_send(method_name)
    result[key] = value unless value.nil?
  rescue ArgumentError, NotImplementedError
    next
  end
end

#fetch_model_detail(_model_name) ⇒ Object

Override in subclasses to make a live API call for model detail. Must return a Hash with symbol keys (e.g. { context_window: 128000 }).



470
471
472
# File 'lib/legion/extensions/llm/provider.rb', line 470

def fetch_model_detail(_model_name)
  nil
end

#find_reachable_url(urls) ⇒ Object



408
409
410
411
412
413
414
# File 'lib/legion/extensions/llm/provider.rb', line 408

def find_reachable_url(urls)
  urls.each do |url|
    full = normalize_url(url)
    return full if url_reachable?(full)
  end
  nil
end

#format_messages(messages) ⇒ Object



318
319
320
321
322
323
324
325
# File 'lib/legion/extensions/llm/provider.rb', line 318

def format_messages(messages)
  messages.map do |msg|
    {
      role: msg.role.to_s,
      content: msg.content
    }
  end
end

#format_tool_calls(_tool_calls) ⇒ Object



327
328
329
# File 'lib/legion/extensions/llm/provider.rb', line 327

def format_tool_calls(_tool_calls)
  nil
end

#headersObject



47
48
49
# File 'lib/legion/extensions/llm/provider.rb', line 47

def headers
  identity_headers
end

#health(live: false) ⇒ Object



160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
# File 'lib/legion/extensions/llm/provider.rb', line 160

def health(live: false)
  readiness_data = readiness(live:)
  raw_health = readiness_data[:health] || readiness_data['health'] || {}
  status = health_status(readiness_data, raw_health)
  {
    provider: slug.to_sym,
    instance_id: provider_instance_id,
    status:,
    ready: readiness_data[:ready] == true || readiness_data['ready'] == true,
    circuit_state: status == 'healthy' ? 'closed' : 'open',
    latency_ms: raw_health[:latency_ms] || raw_health['latency_ms'],
    raw: raw_health
  }.compact
rescue StandardError => e
  handle_exception(e, level: :warn, handled: true, operation: 'llm.provider.health')
  {
    provider: slug.to_sym,
    instance_id: provider_instance_id,
    status: 'unhealthy',
    ready: false,
    circuit_state: 'open',
    error: e.class.name,
    message: e.message
  }
end

#identity_headersObject



51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
# File 'lib/legion/extensions/llm/provider.rb', line 51

def identity_headers
  return {} unless defined?(Legion::Identity::Process) && Legion::Identity::Process.respond_to?(:identity_hash)

  id = Legion::Identity::Process.identity_hash
  hdrs = {
    'x-legion-identity-canonical-name' => id[:canonical_name].to_s,
    'x-legion-identity-trust' => id[:trust].to_s,
    'x-legion-identity-id' => id[:id].to_s,
    'x-legion-identity-kind' => id[:kind].to_s,
    'x-legion-identity-mode' => id[:mode].to_s,
    'x-legion-identity-source' => id[:source].to_s
  }
  hdrs['x-legion-identity-db-principal-id'] = id[:db_principal_id].to_s if id[:db_principal_id]
  hdrs['x-legion-identity-db-identity-id']  = id[:db_identity_id].to_s if id[:db_identity_id]
  hdrs
rescue StandardError
  {}
end

#image(prompt:, model:, size:, with: nil, mask: nil, params: {}) ⇒ Object

rubocop:disable Metrics/ParameterLists



207
208
209
# File 'lib/legion/extensions/llm/provider.rb', line 207

def image(prompt:, model:, size:, with: nil, mask: nil, params: {}) # rubocop:disable Metrics/ParameterLists
  paint(prompt, model:, size:, with:, mask:, params:)
end

#list_models(live: false, **filters) ⇒ Object

rubocop:enable Metrics/ParameterLists



136
137
138
139
140
# File 'lib/legion/extensions/llm/provider.rb', line 136

def list_models(live: false, **filters)
  _ = [live, filters]
  response = @connection.get models_url
  parse_list_models_response response, slug, capabilities
end

#local?Boolean

Returns:

  • (Boolean)


251
252
253
# File 'lib/legion/extensions/llm/provider.rb', line 251

def local?
  self.class.local?
end

#model_allowed?(model_name) ⇒ Boolean

Returns:

  • (Boolean)


366
367
368
369
370
371
372
373
374
375
# File 'lib/legion/extensions/llm/provider.rb', line 366

def model_allowed?(model_name)
  name = model_name.to_s.downcase
  wl = model_whitelist
  bl = model_blacklist

  return false if wl.any? && wl.none? { |p| name.include?(p) }
  return false if bl.any? && bl.any? { |p| name.include?(p) }

  true
end

#model_blacklistObject



344
345
346
347
348
349
# File 'lib/legion/extensions/llm/provider.rb', line 344

def model_blacklist
  bl = config.model_blacklist if config.respond_to?(:model_blacklist)
  bl ||= settings[:model_blacklist] if respond_to?(:settings) && settings.is_a?(Hash)
  bl ||= runtime_provider_setting(:model_blacklist)
  Array(bl).map { |p| p.to_s.downcase }
end

#model_cache_get(key) ⇒ Object



445
446
447
448
449
450
451
452
# File 'lib/legion/extensions/llm/provider.rb', line 445

def model_cache_get(key)
  return nil unless defined?(Legion::Cache)

  cache_local_instance? ? local_cache_get(key) : cache_get(key)
rescue StandardError => e
  handle_exception(e, level: :debug, handled: true, operation: 'llm.provider.model_cache_get', key:)
  nil
end

#model_detail(model_name) ⇒ Object



454
455
456
457
458
459
460
461
462
463
464
465
466
# File 'lib/legion/extensions/llm/provider.rb', line 454

def model_detail(model_name)
  key = model_detail_cache_key(model_name)
  cached = cache_get(key)
  return cached if cached

  result = fetch_model_detail(model_name)
  cache_set(key, result, ttl: 86_400) if result
  result
rescue StandardError => e
  handle_exception(e, level: :warn, handled: true, operation: 'llm.provider.model_detail',
                      model: model_name)
  nil
end

#model_whitelistObject

── Model allow-list / deny-list filtering ────────────────────────



337
338
339
340
341
342
# File 'lib/legion/extensions/llm/provider.rb', line 337

def model_whitelist
  wl = config.model_whitelist if config.respond_to?(:model_whitelist)
  wl ||= settings[:model_whitelist] if respond_to?(:settings) && settings.is_a?(Hash)
  wl ||= runtime_provider_setting(:model_whitelist)
  Array(wl).map { |p| p.to_s.downcase }
end

#moderate(input, model:) ⇒ Object



194
195
196
197
198
# File 'lib/legion/extensions/llm/provider.rb', line 194

def moderate(input, model:)
  payload = render_moderation_payload(input, model:)
  response = @connection.post moderation_url, payload
  parse_moderation_response(response, model:)
end

#nameObject



74
75
76
# File 'lib/legion/extensions/llm/provider.rb', line 74

def name
  self.class.name
end

#normalize_url(url) ⇒ Object



400
401
402
403
404
405
406
# File 'lib/legion/extensions/llm/provider.rb', line 400

def normalize_url(url)
  raw = url.to_s.strip
  return raw if raw.match?(%r{^https?://})

  scheme = tls_enabled? ? 'https' : 'http'
  "#{scheme}://#{raw}"
end

#offering_tierObject



383
384
385
# File 'lib/legion/extensions/llm/provider.rb', line 383

def offering_tier
  config.respond_to?(:tier) ? config.tier : self.class.default_tier
end

#offering_transportObject

── Offering defaults ─────────────────────────────────────────────



379
380
381
# File 'lib/legion/extensions/llm/provider.rb', line 379

def offering_transport
  config.respond_to?(:transport) ? config.transport : self.class.default_transport
end

#paint(prompt, model:, size:, with: nil, mask: nil, params: {}) ⇒ Object

rubocop:disable Metrics/ParameterLists



200
201
202
203
204
205
# File 'lib/legion/extensions/llm/provider.rb', line 200

def paint(prompt, model:, size:, with: nil, mask: nil, params: {}) # rubocop:disable Metrics/ParameterLists
  validate_paint_inputs!(with:, mask:)
  payload = render_image_payload(prompt, model:, size:, with:, mask:, params:)
  response = @connection.post images_url(with:, mask:), payload
  parse_image_response(response, model:)
end

#parse_error(response) ⇒ Object



296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
# File 'lib/legion/extensions/llm/provider.rb', line 296

def parse_error(response)
  return if response.body.empty?

  body = try_parse_json(response.body)
  case body
  when Hash
    error = body['error']
    return error if error.is_a?(String)

    body.dig('error', 'message')
  when Array
    body.map do |part|
      error = part['error']
      error.is_a?(String) ? error : part.dig('error', 'message')
    end.join('. ')
  when String
    body[/"message"\s*:\s*"([^"]{1,500})/, 1] || body
  else
    body
  end
end

#parse_tool_calls(_tool_calls) ⇒ Object



331
332
333
# File 'lib/legion/extensions/llm/provider.rb', line 331

def parse_tool_calls(_tool_calls)
  nil
end

#provider_instance_idObject



484
485
486
487
488
# File 'lib/legion/extensions/llm/provider.rb', line 484

def provider_instance_id
  return config.instance_id.to_sym if config.respond_to?(:instance_id) && config.instance_id

  :default
end

#readiness(live: false) ⇒ Object



263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
# File 'lib/legion/extensions/llm/provider.rb', line 263

def readiness(live: false)
   = {
    provider: slug.to_sym,
    name: name,
    configured: configured?,
    ready: configured?,
    local: local?,
    remote: remote?,
    api_base: api_base,
    endpoints: endpoint_manifest,
    live: live
  }

  return .merge(health: { checked: false }) unless live && [:endpoints][:health]

  response = @connection.get([:endpoints][:health])
  .merge(ready: configured? && health_ready?(response.body), health: response.body)
rescue StandardError => e
  handle_exception(e, level: :warn, handled: true, operation: 'llm.provider.readiness')
  .merge(ready: false, health: { error: e.class.name, message: e.message })
end

#remote?Boolean

Returns:

  • (Boolean)


255
256
257
# File 'lib/legion/extensions/llm/provider.rb', line 255

def remote?
  self.class.remote?
end

#resolve_base_urlObject

── Multi-host base_url resolution ────────────────────────────────



389
390
391
392
393
394
# File 'lib/legion/extensions/llm/provider.rb', line 389

def resolve_base_url
  urls = Array(config_base_url)
  return nil if urls.empty?

  @resolve_base_url ||= find_reachable_url(urls) || normalize_url(urls.first)
end

#runtime_provider_setting(key) ⇒ Object



351
352
353
354
355
356
357
358
359
360
361
362
363
364
# File 'lib/legion/extensions/llm/provider.rb', line 351

def runtime_provider_setting(key)
  return nil unless defined?(Legion::Settings)

  ext = Legion::Settings[:extensions]
  return nil unless ext.is_a?(Hash) && ext[:llm].is_a?(Hash)

  provider_key = self.class.respond_to?(:slug) ? self.class.slug.to_sym : nil
  return nil unless provider_key

  provider_conf = ext[:llm][provider_key]
  provider_conf.is_a?(Hash) ? provider_conf[key] : nil
rescue StandardError
  nil
end

#slugObject



70
71
72
# File 'lib/legion/extensions/llm/provider.rb', line 70

def slug
  self.class.slug
end

#stream_chat(messages:, model:, tools: [], temperature: nil, params: {}, headers: {}, schema: nil, thinking: nil, tool_prefs: nil) ⇒ Object



92
93
94
95
# File 'lib/legion/extensions/llm/provider.rb', line 92

def stream_chat(messages:, model:, tools: [], temperature: nil, params: {}, headers: {}, schema: nil,
                thinking: nil, tool_prefs: nil, &)
  complete(messages, tools:, temperature:, model:, params:, headers:, schema:, thinking:, tool_prefs:, &)
end

#strip_scheme(url) ⇒ Object



416
417
418
# File 'lib/legion/extensions/llm/provider.rb', line 416

def strip_scheme(url)
  url.to_s.sub(%r{^https?://}, '')
end

#tls_enabled?Boolean

Returns:

  • (Boolean)


431
432
433
434
# File 'lib/legion/extensions/llm/provider.rb', line 431

def tls_enabled?
  tls = respond_to?(:settings) ? settings[:tls] : nil
  tls.is_a?(Hash) && tls[:enabled] == true
end

#transcribe(audio_file, model:, language:) ⇒ Object



219
220
221
222
223
224
# File 'lib/legion/extensions/llm/provider.rb', line 219

def transcribe(audio_file, model:, language:, **)
  file_part = build_audio_file_part(audio_file)
  payload = render_transcription_payload(file_part, model:, language:, **)
  response = @connection.post transcription_url, payload
  parse_transcription_response(response, model:)
end

#url_reachable?(url) ⇒ Boolean

Returns:

  • (Boolean)


420
421
422
423
424
425
426
427
428
429
# File 'lib/legion/extensions/llm/provider.rb', line 420

def url_reachable?(url)
  require 'uri'
  require 'socket'
  uri = URI.parse(url)
  Socket.tcp(uri.host, uri.port, connect_timeout: 1).close
  true
rescue StandardError => e
  handle_exception(e, level: :debug, handled: true, operation: 'llm.provider.url_reachable', url:)
  false
end