The daily firehose — everything the toolchain shipped today, already filtered.
// HOW THIS ISSUE IS MADE
We read every release from the 174 tools on our watchlist at the source — GitHub and GitLab release notes, vendor release pages and changelogs, project blogs and feeds, vendor press releases, and the source code behind the tag. Bug-fix-only releases and non-product newsroom noise are dropped; what's left is summarized down to the new capability, how to try it, and any screenshots or videos the release itself published. Every entry links to the sources it was built from.
Agno v2.4.1 adds N1N model provider, Excel knowledge ingestion, and private GitHub/SharePoint file support.
└──▷ GET THIS VERSION
$ git clone --branch v2.4.1 https://github.com/agno-agi/agno.git
# already have the repo? check out this version:$ git checkout v2.4.1
›Adds collect_metrics_on_completion flag to streaming runs, collecting metrics only from the final chunk rather than every streamed chunk.
›Adds tool_call_id to CustomEvent objects yielded from tools, enabling trace events to be linked back to their originating tool calls.
›Adds n1n.ai as a new OpenAI-compatible model provider.
›Adds first-class Excel ingestion (.xlsx/.xls) to Knowledge, parsing workbooks per sheet into separate documents with sheet metadata by routing through the existing CSV reader.
›Adds support for files in private GitHub and SharePoint repositories to be added to Knowledge, available in both the SDK and API.
›Adds node daemon service management with install, status, start, stop, and restart lifecycle commands.
›Adds hybrid BM25 + vector (FTS5) memory search with weighted merging and fallback.
+34 moreshow less
›Adds SQLite embedding cache to accelerate memory reindexing and frequent updates.
›Adds OpenAI batch indexing for embeddings with parallel jobs (default concurrency: 2) and 2-second polling.
›Adds native Gemini embeddings provider for memory search.
›Adds --verbose flag for detailed memory status and batch indexing logs.
›Adds /v1/responses (OpenResponses) API endpoint with item-based input, semantic streaming events, file/image inputs, tool_choice, usage, and output limits.
›Adds /usage cost summaries and macOS menu cost charts.
›Adds the Nostr channel plugin with profile management and onboarding defaults.
›Ships the bundled BlueBubbles channel plugin (disabled by default).
›Adds Zalouser channel with dock metadata, config schema, and status reporting.
›Migrates Matrix integration to matrix-bot-sdk with E2EE support, location handling, and group allowlist upgrades.
›Adds HTTP webhook mode for Slack via Bolt HTTP receiver.
›Enriches Telegram forwarded-message context with normalized origin details and legacy fallback.
›Exposes Discord /skill command globally with fallback when native command limits are exceeded.
›Adds Qwen Portal OAuth provider support.
›Adds plugin slots with a dedicated memory slot selector.
›Adds optional agent tools with explicit allowlists and a plugin tool authoring guide.
›Requires manifest-embedded config schemas with preflight validation warnings for plugins.
›Auto-enables bundled channel/provider plugins when configuration is present.
›Adds daily session reset policy with per-type overrides and idle windows (default 4am local).
›Allows sessions_spawn to override thinking level for sub-agent runs.
›Adds security warning when models ≤300B run without sandboxing while web tools are enabled.
›Adds host/security/ask routing for gateway and node exec via Exec tool.
›Adds /exec directive for per-session exec defaults (host/security/ask/node).
›Migrates exec approvals to ~/.clawdbot/exec-approvals.json with per-agent allowlists, skill auto-allow toggle, approvals UI, and node exec lifecycle events.
›Adds TUI syntax highlighting for code blocks.
›Adds TUI session picker with derived titles, fuzzy search, relative timestamps, and last-message preview.
›Adds searchable model picker in the TUI for faster model selection.
›Adds input history (up/down navigation) for submitted messages in the TUI.
›Adds copy-as-markdown with error feedback to the Control UI.
›Adds skill downloads with OS-filtered options.
›Adds the local sherpa-onnx-tts skill.
›Adds config timestamp stamping and warns when the config file is newer than the running build.
›Migrates Android node transport to the Gateway WebSocket protocol with TLS pinning and gateway discovery naming.
›Syncs plugin sources on channel switches and updates npm-installed plugins during clawdbot update.
└──▷ BREAKING ON UPGRADE
!Invalid or unknown config entries are now rejected and will prevent the gateway from starting; run clawdbot doctor --fix to repair, then clawdbot plugins update if you use plugins.
Qwen Code v0.8.0-preview.0 adds /bug in non-interactive mode, dim YOLO-mode borders, and a redesigned welcome screen.
└──▷ GET THIS VERSION
$ git clone --branch v0.8.0-preview.0 https://github.com/QwenLM/qwen-code.git
# already have the repo? check out this version:$ git checkout v0.8.0-preview.0
└──▷ TRY IT
File a bug report from a CI script or automated pipeline without starting an interactive session.
$ qwen --non-interactive /bug
›Adds /bug command support to non-interactive mode, enabling automated bug-reporting workflows without an interactive session.
›Redesigns the CLI welcome screen and settings dialog with an updated layout.
›Adds dim colors for YOLO/auto-accept mode borders to visually distinguish that operating mode.
›Adds agent.show_turn_stats setting (default false) to display a timer, token count, and active turn indicator in the Agent panel.
›Adds tab_bar.show_pinned_tabs_in_separate_row setting to display pinned and unpinned tabs in separate rows, giving unpinned tabs full horizontal space.
›Adds local parameter to terminal commands to open local terminals inside remote projects.
›Adds match_quotes parameter to the vim::Matching action to control whether % matches quote characters.
›Adds vim::SearchUnderCursor and vim::SearchUnderCursorPrevious actions (bound to * / #) to search under the cursor without immediately jumping to the next or previous match.
+20 moreshow less
›Adds {'ignore_newlines': true} opt-in parameter to editor: delete to previous subword start and editor: delete to next subword end action bindings to restore previous newline behavior.
›Unifies the branch, worktree, and stash pickers into a single Git picker for easier navigation between them.
›Adds structured outline support for Markdown files.
›Adds total and per-file line-added/removed counts for Agent threads.
›Adds editor: move to {start,end} of larger syntax node actions to move the cursor to the start or end of the parent tree-sitter node.
›Adds a project dropdown to the workspace for clearer display of the active project and easier switching between multiple projects.
›Enables the Inline Assistant to display messages and reduces non-code text insertion.
›Adds Copilot support for jumping to unresolved diagnostics in edit predictions.
›Adds keyboard navigation between views in the debugger's new session picker.
›Adds line numbers to the project symbols picker to distinguish between symbols with the same name.
›Adds scrollbar to the outline view picker.
›Adds the ability to right-click a breadcrumb file-path button to copy the path.
›Pressing esc when the Agent panel is focused now interrupts the running thread even if the text input box is not specifically focused.
›Stopping a terminal tool call in the Agent panel now also cancels the thread generation and captures output up to that point so the model can see what was running.
›Improves the Ollama settings page to persist the current URL value when clicking 'Connect'.
›Improves performance of project search in remote projects.
›Adds '#' as part of a word when double-clicking text in .json and .css files.
›Improves the % motion to always fall back to text-based bracket matching when language-aware matching fails.
›Adds ability to customize Vim mode foreground color.
›Improves Windows credential write failures to display the underlying Credential Manager error message instead of a generic error.
Phoenix client 1.28.0 adds FaithfulnessEvaluator and a span_id_key for linking dataset examples to traces.
└──▷ GET THIS VERSION
$ git clone --branch arize-phoenix-client-v1.28.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:$ git checkout arize-phoenix-client-v1.28.0
›Adds span_id_key parameter to link dataset examples to traces.
›Adds FaithfulnessEvaluator class for faithfulness evaluation of LLM outputs.
Phoenix v12.31.0 adds a CLI, FaithfulnessEvaluator, span-to-trace linking, and dataset/experiment CLI commands.
└──▷ GET THIS VERSION
$ git clone --branch arize-phoenix-v12.31.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:$ git checkout arize-phoenix-v12.31.0
└──▷ USE IT
Score LLM responses for faithfulness directly from the Phoenix evaluator library instead of the deprecated HallucinationEvaluator.
python
from phoenix.evals import FaithfulnessEvaluator
evaluator = FaithfulnessEvaluator(model=model)
results = evaluator.evaluate(dataframe=df)
›Adds span_id_key parameter to link dataset examples to traces, enabling traceability between datasets and recorded spans.
›Introduces a phoenix CLI with dataset and experiment subcommands for managing datasets and experiments from the command line.
›Adds FaithfulnessEvaluator, a new built-in LLM classification evaluator for assessing response faithfulness.
$ git clone --branch python-v0.27.0-beta.0 https://github.com/lancedb/lancedb.git
# already have the repo? check out this version:$ git checkout python-v0.27.0-beta.0
›Exposes table URI via a new uri property on table objects, giving callers direct access to the underlying storage path.
›Adds support for remote IVF-RQ (Inverted File with Residual Quantization) indexing, enabling compressed vector search on remote tables.
›Parallelizes embedding computations, reducing latency when generating embeddings for large batches.
›Adds voyage-multimodal-3.5 as a supported embedding model for multimodal (text + image) retrieval.
›Enables the HuggingFace embedding feature by default, removing the need for manual opt-in configuration.
└──▷ BREAKING ON UPGRADE
!The Rust crate's default Cargo features have been removed; any project relying on default features must now explicitly declare the features it needs in its Cargo.toml.
LanceDB v0.24.0-beta.0 adds Voyage multimodal embeddings, remote IVF-RQ index support, parallel embedding computation, and exposes table URIs.
└──▷ GET THIS VERSION
$ git clone --branch v0.24.0-beta.0 https://github.com/lancedb/lancedb.git
# already have the repo? check out this version:$ git checkout v0.24.0-beta.0
›Exposes table URI via a new uri property on table objects, letting callers retrieve the underlying storage path programmatically.
›Adds support for remote IVF-RQ (Inverted File with Residual Quantization) indexing, enabling compressed approximate-nearest-neighbor search on remote tables.
›Adds voyage-multimodal-3.5 as a supported embedding model for multimodal vector generation.
›Parallelizes embedding computations, reducing latency when generating embeddings for large batches.
›Enables the HuggingFace embedding feature by default in the Rust crate, removing the need to opt in via a feature flag.
└──▷ BREAKING ON UPGRADE
!The Rust crate no longer enables default features; any feature previously on by default (other than huggingface, which is now explicitly enabled) must now be opted into explicitly in Cargo.toml.
THE AI TOOLCHAIN