mirror of
https://github.com/we-promise/sure.git
synced 2026-06-04 02:09:01 +00:00
* feat(ai): add Anthropic provider with chat parity (1/5)
Introduces Provider::Anthropic alongside Provider::Openai, implementing
the LlmConcept chat_response contract over the official anthropic Ruby
SDK. Batch ops, PDF, and RAG land in follow-up PRs.
- Provider::Anthropic uses Messages API for sync and streaming responses
- ChatConfig builds requests with ephemeral prompt-cache markers on the
system prompt and the last tool definition
- MessageFormatter reconstructs multi-turn history (text + tool_use +
tool_result blocks) from raw Message records, including the paired
user-role tool_result turn Anthropic requires after every tool_use
- ChatParser maps Anthropic Message into the shared ChatResponse Data
- Registry, Setting, User, Chat default model wired for ANTHROPIC_*
envs and Setting.anthropic_*; LLM_PROVIDER selects between providers
- Responder forwards raw conversation_history (Array<Message>) so
providers without hosted conversation state can rebuild context
- OpenAI provider accepts and ignores the new kwarg (no behavior change)
Tests cover provider init, model gating, MessageFormatter for all turn
shapes, ChatConfig request building (max_tokens, system cache, tool
conversion), ChatParser for text / tool_use / mixed blocks, Registry
discovery, and mocked chat_response success / error / function_request
paths. Live VCR cassettes recorded in a follow-up with a real key.
Stacked PRs: 2/5 batch ops + cost ledger, 3/5 PDF, 4/5 pgvector RAG,
5/5 settings UI + disclosure.
* fix(ai): address PR review on Anthropic provider foundation
Surface fixes raised by Codex + CodeRabbit on PR 1/5:
- Provider::Anthropic#chat_response now accepts (and ignores) a
`messages:` kwarg. Assistant::Responder passes both `messages:`
(OpenAI-shape) and `conversation_history:` (raw Message records) for
cross-provider parity, so the previous signature raised
ArgumentError on the first chat turn through the Anthropic provider.
- Provider::Anthropic#supports_model? bypasses the `claude` prefix
gate when a custom base_url is configured, mirroring the OpenAI
provider. Bedrock-shaped IDs like
`anthropic.claude-sonnet-4-5-20250929-v1:0` and
`claude-opus-4@20250514` are otherwise rejected by
Assistant::Provided#get_model_provider and the chat dies.
- Setting.anthropic_access_token is now in
EncryptedSettingFields::ENCRYPTED_FIELDS so the Anthropic API key
is encrypted at rest like every other provider secret. Previously
plaintext while siblings (openai_access_token, twelve_data_api_key,
external_assistant_token) were ciphertext.
- Chat.default_model falls back to whichever provider is actually
configured. Previously, with LLM_PROVIDER=anthropic but no
Anthropic credentials, the default model resolved to a Claude ID
that no registered provider supported, so chats failed even when
OpenAI was fully configured. Adds Provider::{Anthropic,Openai}#configured?
class methods for the readable callsite.
- Provider::Anthropic.effective_model uses
`ENV["ANTHROPIC_MODEL"].presence || Setting.anthropic_model` so the
Setting lookup is only performed when the env var is absent — the
previous `ENV.fetch(KEY, default)` evaluated the default arg
eagerly on every call.
- Provider::Anthropic::ChatConfig#anthropic_input_schema strips both
`:strict` and `"strict"` keys so JSON-decoded schemas with string
keys cannot leak the OpenAI-only flag through to Anthropic.
Test coverage added: supports_model? bypass on custom endpoints,
chat_response messages: kwarg compatibility, default_model fallback
in the three credential combinations, configured? against ENV +
Setting, strict-flag stripping for both key types, and a
`Setting.expects(:anthropic_model).never` assertion proving the
ENV-precedence test now exercises the lazy path.
All 4365 tests pass (1 pre-existing libvips env error unrelated).
* test(chat): make default_model tests resilient to ENV model overrides
CodeRabbit flagged on PR review: the new default_model tests asserted
against Provider::*::DEFAULT_MODEL, but Chat.default_model actually
returns Provider::*.effective_model.presence (which reads
OPENAI_MODEL / ANTHROPIC_MODEL from the environment). With either env
var set, the tests would fail intermittently even though routing was
correct.
- New default_model tests now assert against the provider's
effective_model directly, so they verify the routing decision
(which provider's value wins) without coupling to the constant.
- Pre-existing "creates with default model" assertions had the same
brittleness; switch them to compare against Chat.default_model so
the chosen model is whatever the env / Setting cascade resolves to.
Verified by running `ANTHROPIC_MODEL=claude-haiku-4-5 OPENAI_MODEL=gpt-4o
bin/rails test test/models/chat_test.rb` — 16 runs, 0 failures
(previously 2 pre-existing failures + 0 from the new tests).
* fix(ai): address local review on Anthropic foundation
- Provider::Anthropic#supports_pdf_processing? bypasses prefix gate for
custom endpoints, mirroring supports_model?
- Provider::Anthropic#initialize raises Error when custom_endpoint? AND
model.blank?, parity with Provider::Openai
- stream_chat_response captures partial usage on mid-stream errors and
records it via the new on_partial callback so chat_response can skip
the duplicate error row in the outer rescue
- safe_accumulated_message swallows the secondary failure when the SDK
cannot reconstruct a snapshot
- langfuse_client memoizes properly (||= instead of =) so repeated calls
don't churn Langfuse instances
- MessageFormatter sorts tool_calls by created_at then id so the
message array is deterministic across replays; skips tool_calls
missing both provider_call_id and provider_id rather than sending
`id: nil` and getting rejected by Anthropic
- Setting.anthropic_access_token default falls back through
ENV["ANTHROPIC_API_KEY"].presence (was missing .presence, so an
empty-string env value bled through)
- User#openai_configured? / #anthropic_configured? delegate to the
Provider::* class methods — single source of truth
- Assistant::Responder renames the OpenAI-shape history builder
conversation_history → openai_messages_payload so the kwarg name
matches the local method name (messages: openai_messages_payload,
conversation_history: chat_message_records)
- Assistant::Builtin stale-history comment updated to reference both
builders
Adds a streaming chat_response test using ad-hoc subclasses of the
SDK event types so the case/when dispatch matches via is_a? without
stubbing class-level === behavior.
* test(ai): add Anthropic tool_use round-trip + multi-tool turn coverage
Addresses @jjmata's "worth confirming" note on PR #1983: tool-use turns
from prior assistant messages must round-trip correctly when retrieved
from the database.
- New `ChatParser → ToolCall::Function → MessageFormatter` test walks
the full path: Anthropic response with a tool_use block →
ChatFunctionRequest → ToolCall::Function.from_function_request →
persisted on the AssistantMessage → MessageFormatter rebuild on the
next turn. Asserts the original `tool_use.id` is preserved end-to-end
as both `tool_use.id` and the paired `tool_result.tool_use_id`, and
that the original `input` hash and serialized result content survive.
- New multi-tool assistant turn test confirms two tool_use blocks on a
single assistant message render as two tool_use blocks followed by
two paired tool_result blocks in a single user-role follow-up,
matching Anthropic's required alternation.
Both tests exercise the existing PR1 code without behavior changes.
* test(ai): require "ostruct" explicitly in Anthropic provider tests
OpenStruct is moving out of Ruby's default load path (warning in 3.4+,
removed in 3.5+). Tests work today because ActiveSupport transitively
loads it, but that's incidental. Match the existing convention in
test/controllers/settings/hostings_controller_test.rb which explicitly
requires ostruct for the same reason.
* fix(ai): sanitize Langfuse warn logs, normalize tool_use.input, dedup history fetch
Addresses three open CodeRabbit findings on PR #1983.
- Provider::Anthropic Langfuse rescue branches no longer include
`e.full_message` in `Rails.logger.warn`. `full_message` bundles the
backtrace + cause chain and on some SDK error types includes the
serialized request/response payload (prompt, model output). Logs
now report `#{e.class}: #{e.message}` only. Three sites:
create_langfuse_trace, log_langfuse_generation, upsert_langfuse_trace.
Note: Provider::Openai has the same pattern (copy-pasted source) —
harmonization deferred to a follow-up cleanup PR; this commit fixes
only the Anthropic provider to keep PR scope tight.
- MessageFormatter#parse_arguments now coerces any non-Hash parsed
result to `{}`. Anthropic's Messages API requires `tool_use.input`
to be a JSON object (map); a stored ToolCall::Function record whose
arguments parse to a scalar, bool, or array (corrupt row, legacy
data, cross-provider bleed) would otherwise produce a payload the
API rejects. Normal flow stores Hash arguments end-to-end so the
fix is defensive — adds 2 tests covering scalar/array JSON strings
and non-String non-Hash inputs.
- Assistant::Responder dedups the chat-history fetch. The previous
layout fired two near-identical `chat.messages.where(...).includes(
:tool_calls).ordered` queries per LLM turn (one for the OpenAI-shape
payload, one for the raw-records kwarg). A new memoized
`complete_chat_messages` fetches once; `chat_message_records` filters
out the current message via `Array#reject`, `openai_messages_payload`
iterates the cached array unchanged. One SQL query per turn instead
of two. Memoization scope = single Responder instance (per LLM call),
so cache invalidation is not a concern.
All 4370 tests pass (1 pre-existing libvips env error unrelated).
Rubocop + brakeman clean.
* fix(ci): replace sk-ant- prefixed test placeholders
Pipelock secret scanner pattern-matches `sk-ant-*` as a real Anthropic
API key and fails the PR security-scan check. Test stubs and
ClimateControl env values used `sk-ant-test`, `sk-ant-from-setting`,
`sk-ant-x`, `sk-ant-y` as obvious placeholders, but the scanner does
not care about value entropy.
Switched to `fake-anthropic-key-*` / `fake-token-*` strings so the
scanner stops flagging them. No production code touched, no behavior
change — Provider::Anthropic still accepts any non-blank token.
288 lines
11 KiB
Ruby
288 lines
11 KiB
Ruby
# Dynamic settings the user can change within the app (helpful for self-hosting)
|
|
class Setting < RailsSettings::Base
|
|
class ValidationError < StandardError; end
|
|
|
|
cache_prefix { "v1" }
|
|
|
|
# Third-party API keys
|
|
field :twelve_data_api_key, type: :string, default: ENV["TWELVE_DATA_API_KEY"]
|
|
field :openai_access_token, type: :string, default: ENV["OPENAI_ACCESS_TOKEN"]
|
|
field :openai_uri_base, type: :string, default: ENV["OPENAI_URI_BASE"]
|
|
field :openai_model, type: :string, default: ENV["OPENAI_MODEL"]
|
|
field :openai_json_mode, type: :string, default: ENV["LLM_JSON_MODE"]
|
|
field :anthropic_access_token, type: :string, default: ENV["ANTHROPIC_ACCESS_TOKEN"].presence || ENV["ANTHROPIC_API_KEY"].presence
|
|
field :anthropic_model, type: :string, default: ENV["ANTHROPIC_MODEL"]
|
|
field :anthropic_base_url, type: :string, default: ENV["ANTHROPIC_BASE_URL"]
|
|
field :llm_provider, type: :string, default: ENV.fetch("LLM_PROVIDER", "openai")
|
|
|
|
# LLM token budget (applies to every outbound LLM call: chat, auto-categorize,
|
|
# merchant detection, enhance-merchants, PDF processing). Defaults track
|
|
# Ollama's historical 2048-token baseline so local small-context models work
|
|
# out of the box. ENV overrides Setting at read time in Provider::Openai.
|
|
field :llm_context_window, type: :integer, default: ENV["LLM_CONTEXT_WINDOW"]&.to_i
|
|
field :llm_max_response_tokens, type: :integer, default: ENV["LLM_MAX_RESPONSE_TOKENS"]&.to_i
|
|
field :llm_max_items_per_call, type: :integer, default: ENV["LLM_MAX_ITEMS_PER_CALL"]&.to_i
|
|
field :external_assistant_url, type: :string
|
|
field :external_assistant_token, type: :string
|
|
field :external_assistant_agent_id, type: :string
|
|
field :brand_fetch_client_id, type: :string, default: ENV["BRAND_FETCH_CLIENT_ID"]
|
|
field :brand_fetch_high_res_logos, type: :boolean, default: ENV.fetch("BRAND_FETCH_HIGH_RES_LOGOS", "false") == "true"
|
|
|
|
BRAND_FETCH_LOGO_SIZE_STANDARD = 40
|
|
BRAND_FETCH_LOGO_SIZE_HIGH_RES = 120
|
|
# Matches both legacy single-segment URLs (`/apple.com/icon/...`) and
|
|
# explicit type-routed URLs introduced 2026 (`/crypto/BTC/icon/...`,
|
|
# `/domain/apple.com/icon/...`). `[^?]+` reaches across the extra slash
|
|
# so transform_brand_fetch_url can rewrite the size params on both shapes.
|
|
BRAND_FETCH_URL_PATTERN = %r{(https://cdn\.brandfetch\.io/[^?]+/icon/fallback/lettermark/)w/\d+/h/\d+(\?c=.+)}
|
|
|
|
def self.brand_fetch_logo_size
|
|
brand_fetch_high_res_logos ? BRAND_FETCH_LOGO_SIZE_HIGH_RES : BRAND_FETCH_LOGO_SIZE_STANDARD
|
|
end
|
|
|
|
# Transforms a stored Brandfetch URL to use the current logo size setting
|
|
def self.transform_brand_fetch_url(url)
|
|
return url unless url.present? && url.match?(BRAND_FETCH_URL_PATTERN)
|
|
|
|
size = brand_fetch_logo_size
|
|
url.gsub(BRAND_FETCH_URL_PATTERN, "\\1w/#{size}/h/#{size}\\2")
|
|
end
|
|
|
|
# Provider selection
|
|
field :exchange_rate_provider, type: :string, default: ENV.fetch("EXCHANGE_RATE_PROVIDER", "twelve_data")
|
|
field :securities_provider, type: :string, default: ENV.fetch("SECURITIES_PROVIDER", "twelve_data")
|
|
|
|
# Multi-provider: comma-separated list of enabled securities providers
|
|
field :securities_providers, type: :string, default: ENV.fetch("SECURITIES_PROVIDERS", "")
|
|
|
|
# New provider API keys (encrypted at rest — see EncryptedSettingFields below)
|
|
field :tiingo_api_key, type: :string, default: ENV["TIINGO_API_KEY"]
|
|
field :eodhd_api_key, type: :string, default: ENV["EODHD_API_KEY"]
|
|
field :alpha_vantage_api_key, type: :string, default: ENV["ALPHA_VANTAGE_API_KEY"]
|
|
|
|
# Transparent encryption for API key fields. The `field` macro defines the
|
|
# raw getter/setter on the class. By prepending this module we intercept
|
|
# reads (decrypt) and writes (encrypt) while `super` delegates to the
|
|
# original getter/setter generated by rails-settings-cached.
|
|
#
|
|
# Backward-compatible: if decryption fails (e.g. the value was stored before
|
|
# encryption was enabled) the raw value is returned as-is.
|
|
module EncryptedSettingFields
|
|
ENCRYPTED_FIELDS = %i[
|
|
twelve_data_api_key
|
|
tiingo_api_key
|
|
eodhd_api_key
|
|
alpha_vantage_api_key
|
|
openai_access_token
|
|
anthropic_access_token
|
|
external_assistant_token
|
|
].freeze
|
|
|
|
ENCRYPTED_FIELDS.each do |field_name|
|
|
define_method(field_name) do
|
|
raw = super()
|
|
decrypt_setting(raw)
|
|
end
|
|
|
|
define_method(:"#{field_name}=") do |value|
|
|
super(encrypt_setting(value))
|
|
end
|
|
end
|
|
|
|
private
|
|
|
|
def setting_encryptor
|
|
@setting_encryptor ||= begin
|
|
key = ActiveSupport::KeyGenerator.new(
|
|
Rails.application.secret_key_base
|
|
).generate_key("setting_encryption", 32)
|
|
ActiveSupport::MessageEncryptor.new(key)
|
|
end
|
|
end
|
|
|
|
def encrypt_setting(value)
|
|
return value if value.blank?
|
|
setting_encryptor.encrypt_and_sign(value)
|
|
end
|
|
|
|
def decrypt_setting(value)
|
|
return value if value.blank?
|
|
setting_encryptor.decrypt_and_verify(value)
|
|
rescue ActiveSupport::MessageVerifier::InvalidSignature,
|
|
ActiveSupport::MessageEncryptor::InvalidMessage
|
|
# Value was stored before encryption was enabled — return as-is.
|
|
# It will be re-encrypted on next write.
|
|
value
|
|
end
|
|
end
|
|
|
|
class << self
|
|
prepend EncryptedSettingFields
|
|
end
|
|
|
|
def self.enabled_securities_providers
|
|
plural = ENV["SECURITIES_PROVIDERS"].presence || securities_providers.presence
|
|
if plural.present?
|
|
plural.to_s.split(",").map(&:strip).reject(&:blank?)
|
|
else
|
|
# Backward compat: fall back to singular setting
|
|
[ ENV["SECURITIES_PROVIDER"].presence || securities_provider ].compact
|
|
end
|
|
end
|
|
|
|
# Sync settings - check both provider env vars for default
|
|
# Only defaults to true if neither provider explicitly disables pending
|
|
SYNCS_INCLUDE_PENDING_DEFAULT = begin
|
|
simplefin = ENV.fetch("SIMPLEFIN_INCLUDE_PENDING", "1") == "1"
|
|
plaid = ENV.fetch("PLAID_INCLUDE_PENDING", "1") == "1"
|
|
simplefin && plaid
|
|
end
|
|
field :syncs_include_pending, type: :boolean, default: SYNCS_INCLUDE_PENDING_DEFAULT
|
|
field :auto_sync_enabled, type: :boolean, default: ENV.fetch("AUTO_SYNC_ENABLED", "1") == "1"
|
|
field :auto_sync_time, type: :string, default: ENV.fetch("AUTO_SYNC_TIME", "02:22")
|
|
field :auto_sync_timezone, type: :string, default: ENV.fetch("AUTO_SYNC_TIMEZONE", "UTC")
|
|
|
|
AUTO_SYNC_TIME_FORMAT = /\A([01]?\d|2[0-3]):([0-5]\d)\z/
|
|
|
|
def self.valid_auto_sync_time?(time_str)
|
|
return false if time_str.blank?
|
|
AUTO_SYNC_TIME_FORMAT.match?(time_str.to_s.strip)
|
|
end
|
|
|
|
def self.valid_auto_sync_timezone?(timezone_str)
|
|
return false if timezone_str.blank?
|
|
ActiveSupport::TimeZone[timezone_str].present?
|
|
end
|
|
|
|
# Dynamic fields are now stored as individual entries with "dynamic:" prefix
|
|
# This prevents race conditions and ensures each field is independently managed
|
|
|
|
# Onboarding and app settings
|
|
ONBOARDING_STATES = %w[open closed invite_only].freeze
|
|
DEFAULT_ONBOARDING_STATE = begin
|
|
env_value = ENV["ONBOARDING_STATE"].to_s.presence || "open"
|
|
ONBOARDING_STATES.include?(env_value) ? env_value : "open"
|
|
end
|
|
|
|
field :onboarding_state, type: :string, default: DEFAULT_ONBOARDING_STATE
|
|
field :require_invite_for_signup, type: :boolean, default: false
|
|
field :require_email_confirmation, type: :boolean, default: ENV.fetch("REQUIRE_EMAIL_CONFIRMATION", "true") == "true"
|
|
field :invite_only_default_family_id, type: :string, default: nil
|
|
|
|
def self.validate_onboarding_state!(state)
|
|
return if ONBOARDING_STATES.include?(state)
|
|
|
|
raise ValidationError, I18n.t("settings.hostings.update.invalid_onboarding_state")
|
|
end
|
|
|
|
class << self
|
|
alias_method :raw_onboarding_state, :onboarding_state
|
|
alias_method :raw_onboarding_state=, :onboarding_state=
|
|
alias_method :raw_openai_model, :openai_model
|
|
alias_method :raw_openai_model=, :openai_model=
|
|
|
|
def onboarding_state
|
|
value = raw_onboarding_state
|
|
return "invite_only" if value.blank? && require_invite_for_signup
|
|
|
|
value.presence || DEFAULT_ONBOARDING_STATE
|
|
end
|
|
|
|
def onboarding_state=(state)
|
|
validate_onboarding_state!(state)
|
|
self.require_invite_for_signup = state == "invite_only"
|
|
self.raw_onboarding_state = state
|
|
end
|
|
|
|
def openai_model=(value)
|
|
old_value = raw_openai_model
|
|
self.raw_openai_model = value
|
|
|
|
if old_value != value && old_value.present?
|
|
Rails.logger.info("OpenAI model changed from #{old_value} to #{value}, clearing AI cache for all families")
|
|
Family.find_each do |family|
|
|
ClearAiCacheJob.perform_later(family)
|
|
end
|
|
end
|
|
end
|
|
|
|
# Support dynamic field access via bracket notation
|
|
# First checks if it's a declared field, then falls back to individual dynamic entries
|
|
def [](key)
|
|
key_str = key.to_s
|
|
|
|
# Check if it's a declared field first
|
|
if respond_to?(key_str)
|
|
public_send(key_str)
|
|
else
|
|
# Fall back to individual dynamic entry lookup
|
|
find_by(var: dynamic_key_name(key_str))&.value
|
|
end
|
|
end
|
|
|
|
def []=(key, value)
|
|
key_str = key.to_s
|
|
|
|
# If it's a declared field, use the setter
|
|
if respond_to?("#{key_str}=")
|
|
public_send("#{key_str}=", value)
|
|
else
|
|
# Store as individual dynamic entry
|
|
dynamic_key = dynamic_key_name(key_str)
|
|
if value.nil?
|
|
where(var: dynamic_key).destroy_all
|
|
clear_cache
|
|
else
|
|
# Use upsert for atomic insert/update to avoid race conditions
|
|
upsert({ var: dynamic_key, value: value.to_yaml }, unique_by: :var)
|
|
clear_cache
|
|
end
|
|
end
|
|
end
|
|
|
|
# Check if a dynamic field exists (useful to distinguish nil value vs missing key)
|
|
def key?(key)
|
|
key_str = key.to_s
|
|
return true if respond_to?(key_str)
|
|
|
|
# Check if dynamic entry exists
|
|
where(var: dynamic_key_name(key_str)).exists?
|
|
end
|
|
|
|
# Delete a dynamic field
|
|
def delete(key)
|
|
key_str = key.to_s
|
|
return nil if respond_to?(key_str) # Can't delete declared fields
|
|
|
|
dynamic_key = dynamic_key_name(key_str)
|
|
value = self[key_str]
|
|
where(var: dynamic_key).destroy_all
|
|
clear_cache
|
|
value
|
|
end
|
|
|
|
# List all dynamic field keys (excludes declared fields)
|
|
def dynamic_keys
|
|
where("var LIKE ?", "dynamic:%").pluck(:var).map { |var| var.sub(/^dynamic:/, "") }
|
|
end
|
|
|
|
private
|
|
|
|
def dynamic_key_name(key_str)
|
|
"dynamic:#{key_str}"
|
|
end
|
|
end
|
|
|
|
# Validates OpenAI configuration requires model when custom URI base is set
|
|
def self.validate_openai_config!(uri_base: nil, model: nil)
|
|
# Use provided values or current settings
|
|
uri_base_value = uri_base.nil? ? openai_uri_base : uri_base
|
|
model_value = model.nil? ? openai_model : model
|
|
|
|
# If custom URI base is set, model must also be set
|
|
if uri_base_value.present? && model_value.blank?
|
|
raise ValidationError, "OpenAI model is required when custom URI base is configured"
|
|
end
|
|
end
|
|
end
|