1. Docs
  2. Changes and releases
  3. Developer Portal

Developer Portal

  • Developer Portal backstage v0.129.3

    Fixed

    • Stabilise the AI chat transport across React renders. useChatSetup constructed a fresh AssistantChatTransport on every render and handed it to useChatRuntime; when its identity changed mid-stream (e.g. as a side effect of a state update triggered by streamed reasoning-delta / tool-input-delta events), the runtime tore down the in-flight chat fetch with TypeError: network error. Envoy logged the symptom as response_flags: DC (downstream remote disconnect) against a healthy Backstage upstream while the SSE stream summary was sawFinish=false, last event tool-input-delta, surfacing in the UI as a “Network error” banner even though no real network outage occurred. The transport is now memoised on the resolved API URL and the stable getHeaders / debugFetch callbacks, so a single transport lives for the component’s lifetime.
    • AI chat instrumentation: observe the caller’s AbortSignal and log abort events with the reason inline (ABORT signaled by client at <ms> -- reason: <name>: <message>). Stream-outcome lines now annotate aborted streams with [client-aborted at Nms reason="..."] so a “Network error” banner can be classified as a deliberate client cancel (transport rebuild, unmount, manual stop) versus a real proxy / network failure (no abort signal fired, raw stream error). Read errors that fired after the caller aborted are now logged as console.warn (“cancelled by client AbortSignal”) rather than console.error (“STREAM READ FAILED”), reserving the latter for the genuine pre-finish, non-aborted failure mode.

    See ./docs/releases/v0.129.3-changelog.md for more information.

  • Developer Portal backstage v0.129.2

    Fixed

    • AI chat instrumentation: render the SSE stream summary inline in the log message (bytes=... events=... sawFinish=... lastEventType=...) instead of attaching it as a trailing object argument, so it stays readable in browser-console consumers that flatten the args array (devtools-snapshotters, log shippers, the Cursor IDE browser, etc.) rather than collapsing to “[object Object]”. Also classify a stream read error that arrives after a finish event was already parsed as a post-completion teardown (console.warn, message already committed) instead of a network failure (console.error); this matches the actual semantics of the AI SDK aborting the underlying fetch once it has finished consuming the stream. See ./docs/releases/v0.129.2-changelog.md for more information.
  • Developer Portal backstage v0.129.1

    Fixed

    • AI chat: always log network-level diagnostics (request URL, response status / headers, fetch errors, SSE stream lifecycle including premature termination) to the browser console so a “Network error” banner can be triaged from a user report without requiring the ai-chat-verbose-debugging feature flag. Verbose payload-level logging (messages, system prompt, tool schemas, per-event SSE detail) remains gated on the feature flag in non-production builds. See ./docs/releases/v0.129.1-changelog.md for more information.
  • Developer Portal backstage v0.129.0

    Added

    • Make AI chat sampling parameters configurable per installation. The plugin previously called streamText() without temperature, topP, topK, seed, minP, or maxOutputTokens, so the server’s defaults applied – which for vLLM means temperature=1.0, top_p=1.0, top_k=-1, seed=null. That is far too loose for a tool-using agent backed by a reasoning model and was the dominant cause of token-cost variance in production agent loops (same prompt, fresh chat, observed total-token spread of 22k / 607k / 22k across three runs against the same Qwen3 endpoint). Config now accepts an aiChat.sampling block with temperature, topP, topK, minP, seed, and maxOutputTokens; all fields are optional and default behaviour with no sampling: block is unchanged. temperature, topP, topK, seed, and maxOutputTokens are forwarded through the AI SDK to every provider that supports them. minP is spliced into the request body via the OpenAI-compatible provider’s transformRequestBody hook, since vLLM accepts it as a top-level field but it is not part of the AI SDK call settings. The ai-chat-backend README documents recommended values per model family (Qwen3 thinking/non-thinking, Qwen3-Coder, GPT-4 / GPT-4o, Anthropic Claude). See ./docs/releases/v0.129.0-changelog.md for more information.
  • Developer Portal backstage v0.128.1

    Fixed

    • Fix AI chat backend crash on every chat send. The tools field in the request body is optional, so it arrives as undefined whenever the frontend does not register any client-side tools. The previous implementation called Object.entries(tools) unconditionally, which threw TypeError: Cannot convert undefined or null to object, returning a 500 from POST /api/ai-chat/chat on every request. The browser surfaced this as “network error” and the LLM never reached the tool-merge step, so MCP-provided tools (muster, prometheus, kubernetes, …), skill tools, getDate, and the context-usage tool were also unreachable from chat. frontendTools now accepts null/undefined and treats them as an empty registry.
    • Rebuild MCP clients when the underlying transport closes. The AI chat backend caches one MCPClient per server for 30 minutes, but the cache was holding on to clients whose StreamableHTTP transport had already been torn down by muster (idle timeout, server reset, …). Tool calls then failed with MCPClientError: Attempted to send a request from a closed client until the TTL expired, surfacing in the browser as a “network error” banner. The cache now chains into the transport’s onclose callback and evicts the entry as soon as the connection drops, so the next request rebuilds the client cleanly. Tool execution also detects the same SDK error eagerly and marks the entry dead so a single failure – not a 30-minute window – triggers reconnection. See ./docs/releases/v0.128.1-changelog.md for more information.
  • Developer Portal backstage v0.128.0

    Added

    • Add “Configure with AI” button to App Deployment template.

    Fixed

    • Render AI chat user messages as plain text with preserved whitespace so line breaks (shift+enter) display consistently instead of being collapsed by markdown rendering. See ./docs/releases/v0.128.0-changelog.md for more information.
  • Developer Portal backstage v0.127.0

    Added

    • Allow customization of theme colors via app.branding.theme configuration.
    • Make AI-chat skills configurable per deployment via aiChat.skills (opt out of bundled skills with bundled: false, load extra skills from a directory with dir, or define them inline with inline).

    Fixed

    • Make the sqlite-persistence CI test case (ci-values-case3-sqlite-persistent.yaml) inject a real backend.database.connection.directory override and update the surrounding comment so it reflects what current Backstage actually accepts. The previous comment suggested connection: /var/lib/backstage/db.sqlite, which the Sqlite3Connector rejects at runtime with "connection.filename" is not supported for the base sqlite connection. See ./docs/releases/v0.127.0-changelog.md for more information.
  • Developer Portal backstage v0.126.0

    Added

    • Added configuration options to allow for persisting the SQLite database.
    • Add Mermaid diagram support to AI chat markdown rendering.
    • Render <details>/<summary> HTML in AI chat markdown.
    • Allow AI chat MCP servers to act as the logged-in Backstage user via the new useBackstageUserToken option, so user-context tools like auth.who-am-i work without a static external-access token.

    Changed

    • Strip reasoning content from past AI chat assistant messages to reclaim context tokens for Claude conversations.
    • Replace the fa-kubernetes Font Awesome icon with an inline SVG and remove the Font Awesome kit integration.
    • Replace @terasky/backstage-plugin-catalog-mcp-backend with the built-in catalog.query-catalog-entities action from @backstage/plugin-mcp-actions-backend.

    Fixed

    • Fix AI chat input freezing after typing dead keys (e.g., backtick on German keyboard).

    Removed

    • Remove the custom getCurrentUserInfo AI chat agent tool, superseded by the upstream auth.who-am-i tool from the mcp-actions MCP server. See ./docs/releases/v0.126.0-changelog.md for more information.
  • Developer Portal backstage v0.125.0

    Added

    • Persist AI chat conversations and add a conversation history page with rename, delete, batch delete, and pagination, plus a history view in the chat drawer.
    • Add a getDate tool to the AI chat agent for retrieving the current date and time.
    • Prune older tool outputs in AI chat once cumulative tool I/O exceeds a configurable budget (aiChat.pruning.reservedTokens, aiChat.pruning.minimumSavingsTokens).

    Changed

    • Report AI chat context size, output tokens, and cost by requesting token usage from OpenAI-compatible streaming responses (vLLM/KServe).
    • Clarify tool use in the muster system prompt to avoid calling internal tools via muster’s call_tool.

    Fixed

    • Re-observe the element when the ref target changes in useContainerDimensions.
    • Reject empty titles in PATCH /conversations/:id/title with HTTP 400. See ./docs/releases/v0.125.0-changelog.md for more information.
  • Developer Portal backstage v0.124.4

    Added

    • Make sidebar logo image height configurable via app.branding.logo.height.

    Fixed