github maximhq/bifrost ent-v2.2.0-base
Enterprise v2.2.0 base
on GitHub

4 hours ago

What's Changed

  • docs: add MCP auth warning to 2.1.0 changelogs by @impoiler in #7022
  • fix(http): add Cache-Control: no-store default for /api/ responses to prevent CDN caching by @impoiler in #7014
  • feat: add entity_dac per-entity DAC overrides to governance roles in Helm chart by @BearTS in #7024
  • docs: remove resolved MCP auth warning from v2.1.0 changelog by @impoiler in #7026
  • v2.1.1 by @akshaydeo in #7029
  • fixes package.json for integration test for 7day check by @akshaydeo in #7030
  • enterprise docs by @akshaydeo in #7036
  • migration test fixes by @akshaydeo in #7038
  • fixes reserved namespace redaction for bedrock by @akshaydeo in #7039
  • docs: add trusted networks step and troubleshooting for private IdP OIDC discovery by @BearTS in #7040
  • docs: update virtual key API configuration examples by @R-droid101 in #7043
  • Revert "vk allowed models * handling for governance" by @Pratham-Mishra04 in #7053
  • prompt caching docs; and edge out of alpha by @akshaydeo in #7056
  • docs update by @akshaydeo in #7058
  • fix: responses event action string by @TejasGhatte in #7060
  • feat: route prompt cache reloads through the server so enterprise can gossip them by @impoiler in #7061
  • fix(helm): support multiple access profiles on governance roles by @CarlosLanderas in #7044
  • feat: add trusted_networks to Helm and config schema by @BearTS in #7081
  • fix: route runtime responses compat models to responses api by @TejasGhatte in #7071
  • fix: config flag to route to openai compat api in bedrock by @TejasGhatte in #7073
  • fix: dynamic mantle base path and gpt 6 reasoning helpers by @TejasGhatte in #7077
  • fix: bedrock document source by @TejasGhatte in #7079
  • tests: harness test coverage for bedrock by @TejasGhatte in #7083
  • helm: update openai endpoints key field by @TejasGhatte in #7088
  • fix: guardrail identifier in runtime headers by @TejasGhatte in #7095
  • fix: strip Anthropic server tools on Fireworks/vLLM/SGLang Anthropic-compatible endpoints and report drops via DroppedUnsupportedTools by @Pratham-Mishra04 in #7090
  • docs: add use_anthropic_endpoints option for Fireworks Anthropic-compatible Messages endpoint by @Pratham-Mishra04 in #7091
  • docs: add use_anthropic_endpoints option and Anthropic-compatible endpoint routing to SGL provider docs by @Pratham-Mishra04 in #7092
  • docs: document vLLM server tool dropping behavior and dropped_unsupported_tools response field by @Pratham-Mishra04 in #7094
  • feat(configstore): add time-of-day peak/off-peak pricing columns and migration by @impoiler in #6574
  • feat(cost): add time-of-day peak/off-peak pricing with cached timezone resolution and tests by @impoiler in #6575
  • feat(pricing): add off_peak_cost_multiplier and peak_hours fields to PricingPatch schema and patch logic by @impoiler in #6576
  • refactor(ui): extract FormState, defaultFormState, and buildPatchFromForm from sheet into pricingFields.ts by @impoiler in #6577
  • feat(ui): add off_peak_cost_multiplier and peak_hours support to pricing overrides by @impoiler in #6578
  • docs: add time-of-day pricing docs for model catalog, custom pricing, and DeepSeek caveats by @impoiler in #6579
  • test(e2e): add time-of-day pricing Postman collection and Newman runner by @impoiler in #7054
  • docs(logging): clarify CountRecalcTargets matview staleness and Total approximation by @impoiler in #7078
  • feat: add GA realtime transcription support by @danpiths in #7089
  • fixes gpt-oss mistag by @akshaydeo in #7075
  • fixes bedrock mantle streaming missing usage by @akshaydeo in #7076
  • namespace tool name support across providers by @akshaydeo in #7082
  • fix: prevent governance cleanup dump race by @Constantine3 in #7100
  • move tool namespace collision check to datasheet by @akshaydeo in #7084
  • adds custom tripper to fasthttp streaming clients by @akshaydeo in #7104
  • [fix]: Gemini provider - preserve inline image and audio data in chat completions by @Atharva-Kanherkar in #7033
  • adds clickhouse delte rows improvements by @akshaydeo in #7103
  • feat(redaction): surface literal regions and service URLs in plaintext via RedactedIfSecret by @impoiler in #7085
  • feat(mcp-usage-guide): add OAuth and IdP token auth methods alongside virtual key by @impoiler in #7111
  • fix(sidebar): truncate long item titles to prevent overflow by @impoiler in #7069
  • feat(schemas): add RE2 pattern lists for model allow/block rules on Key and ProviderPermit by @impoiler in #6987
  • feat(models): add RE2 pattern twins for allow/block lists on provider keys and virtual key provider configs by @impoiler in #6988
  • feat(ui): add regex pattern support for model allow/block lists via ModelAccessSelector by @impoiler in #6989
  • docs: add allowed_models_patterns and blacklisted_models_patterns RE2 pattern fields to virtual keys, projects, and provider keys by @impoiler in #6990
  • test(governance): add e2e coverage for regex entries in VK model allow/block lists by @impoiler in #7031
  • improve color pallete of graphs by @akshaydeo in #7113
  • fixes responses sse item null regression by @ReStranger in #6395
  • scope fasthttp stale-connection retries to pooled sockets and end retry backoff on cancel by @akshaydeo in #7105
  • cancel in-flight requests when the client closes its socket before the first byte by @akshaydeo in #7106
  • fix(bedrock): assign unique names to untitled document blocks by @Huang-404-Q in #7027
  • fix(core): clear provider response headers on fallback boundaries by @Huang-404-Q in #7021
  • fixes missing stream termination for custom providers by @akshaydeo in #7115
  • fix: handle incomplete fallback errors by @Constantine3 in #7110
  • image & video model fallbacks fixes by @akshaydeo in #7118
  • [fix]: route opencode-zen Responses calls through /v1/chat/completions by @miguelchico in #6819
  • tool_search in invoke flow by @akshaydeo in #6900
  • feat(bedrock): serve Anthropic tool search on Claude via InvokeModel routing by @akshaydeo in #6908
  • fix: filter credential-bearing provider response headers by classifier by @Atharva-Kanherkar in #7121
  • fixes caller cancellations detections by @akshaydeo in #7116
  • oss changes for support for tool calls args redaction in guardrails by @Madhuvod in #6977
  • docs: docs for singulr ai guardrail provider by @Madhuvod in #7055
  • oss: passthru redaction for transformed output guardrails like model armor, bedrock by @Madhuvod in #7049
  • fix(complexity): preserve input for continuation fallback classification when session state is absent by @Madhuvod in #7122
  • fix: max tokens for deepseek provider by @TejasGhatte in #7131
  • revert: withdraw the model pattern fields (#6987 to #7031) by @impoiler in #7133
  • feat(governance): add regex: pattern support to model allow/block lists by @impoiler in #7134
  • fix: omit client_secret in refresh token grant when client has no secret by @Pratham-Mishra04 in #7042
  • chore: fixes framework tests by @sammaji in #7139
  • chore: fixes core tests by @sammaji in #7146
  • fixes azure-firework model output token capping fix by @akshaydeo in #7142
  • Attribute endpoint MCP inspections in the standard logging pipeline by @akshaydeo in #6959
  • turn reasoning context support by @akshaydeo in #7140
  • fix: Helm chart - map perUserHeaderKeys into rendered config.json by @CallumWayve in #6034
  • feat: add normalized error_type label to bifrost_error_requests_total by @roroghost17 in #7141
  • feat(logs): add pinnable search mode dropdown (auto / content / request ID) by @impoiler in #7149
  • chore: helm version 2.1.42 with mcp per user header key support by @BearTS in #7148
  • chore: harness test fixes by @TejasGhatte in #7147
  • fix: bedrock image block in tool result by @TejasGhatte in #7150
  • mid-conversation switch cache checkpoint by @akshaydeo in #7145
  • feat: adds Claude Desktop marketplace support by @danpiths in #7152
  • docs: add Claude Desktop marketplace setup guide by @danpiths in #7153
  • feat(logs): record governance entity names on MCP tool logs by @impoiler in #7154
  • fix/preserve-per-part-media-resolution by @akshaydeo in #7156
  • fix: sort MCP tool schema properties and $defs on conversion by @dougcalobrisi in #7170
  • fix: image edit check in genai by @TejasGhatte in #7173
  • namespace flattening fix for harness tools by @akshaydeo in #7161
  • fix(anthropic): carry server-side tool search through the response path by @akshaydeo in #7162
  • fix(bedrock): keep tool search alive across the invoke ingress (#7155) by @akshaydeo in #7163
  • fix(bedrock): return tool search as server_tool_use on the invoke ingress by @akshaydeo in #7164
  • fix(bedrock): stop streaming tool search as a client tool_use by @akshaydeo in #7165
  • fix(bedrock): keep replayed tool-search blocks across the invoke ingress by @akshaydeo in #7166
  • fixed mcp logs app icon by @akshaydeo in #7167
  • feat: reference Virtual MCPs by name on access profiles and projects by @roroghost17 in #7181
  • fixes: dropped sse chunks in raw response by @akshaydeo in #7184
  • wait_for_usage flag for custom providers by @akshaydeo in #7187
  • fix(gemini): keep generationConfig ExtraParams intact across retries by @VictorRequenaMaisa in #7138
  • [fix]: Resolve vLLM aliases during key selection by @Constantine3 in #6956
  • test fixes by @akshaydeo in #7192
  • changelogs by @akshaydeo in #7194
  • tests: langchain test fixes by @TejasGhatte in #7193
  • adds make recipie to run integration tests locally by @akshaydeo in #7196
  • chore: uv package by @TejasGhatte in #7197
  • fix: cohere integration response conversion by @TejasGhatte in #7198
  • v2.2.0 by @akshaydeo in #7199
  • fixes race condition in skill serving handler by @akshaydeo in #7200

New Contributors

Full Changelog: ent-v2.1.1-base...ent-v2.2.0-base

Check out latest releases or
releases around maximhq/bifrost ent-v2.2.0-base

Don't miss a new bifrost release

NewReleases is sending notifications on new releases.

Get notifications