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 v2.5.3 adds remote S3 knowledge endpoints, PDF content sanitization, and OpenTelemetry extras.
└──▷ GET THIS VERSION
$ git clone --branch v2.5.3 https://github.com/agno-agi/agno.git
# already have the repo? check out this version:$ git checkout v2.5.3
└──▷ USE IT
Disable PDF sanitization when extracting structured content like code blocks or tables where whitespace is semantically significant.
python
from agno.document.reader.pdf import PDFReader
reader = PDFReader(sanitize_content=False)
docs = reader.read("report_with_tables.pdf")
›Adds sanitize_content parameter to BasePDFReader (enabled by default) to normalize fragmented PDF text extraction — collapses word-per-line artifacts while preserving paragraph breaks; set sanitize_content=False to preserve structured content like code or tables.
›Adds API endpoints for listing remote knowledge contents and enables uploading content in S3 buckets via AgentOS.
›Adds is_component, current_version, and stage fields to list endpoints.
›Adds OpenTelemetry and Agno instrumentation dependencies to the os extras.
LangGraph Python SDK 0.3.8 adds stream_mode, stream_subgraphs, stream_resumable, and durability options to cron jobs, plus improved store auth type safety.
└──▷ GET THIS VERSION
$ git clone --branch sdk==0.3.8 https://github.com/langchain-ai/langgraph.git
# already have the repo? check out this version:$ git checkout sdk==0.3.8
›Adds stream_mode, stream_subgraphs, stream_resumable, and durability parameters to cron job creation in the Python SDK.
›Improves type safety and docstrings for store auth in the Python SDK.
OpenAI Agents SDK v0.9.2 adds reasoning_item_id_policy to RunConfig to suppress 400 errors with reasoning models.
└──▷ GET THIS VERSION
$ git clone --branch v0.9.2 https://github.com/openai/openai-agents-python.git
# already have the repo? check out this version:$ git checkout v0.9.2
└──▷ USE IT
Prevent 400 errors when running a reasoning model by opting in to omitting reasoning item IDs across a full multi-turn session.
python
run_config = RunConfig(reasoning_item_id_policy="omit")
result = await Runner.run(
agent,
"Tell me about recursion in programming.",
run_config=run_config,
)
›Adds reasoning_item_id_policy='omit' option to RunConfig to drop reasoning item IDs when using reasoning models, preventing 400 errors from inconsistent item sets; opt-in with default behavior unchanged.
›Persists reasoning_item_id_policy across agent resumes and streamed follow-up turns.
OpenClaw v2026.2.19 adds Apple Watch support, APNs push delivery, paired-device hygiene CLI commands, and gateway HTTP no-auth audit findings.
└──▷ GET THIS VERSION
$ git clone --branch v2026.2.19 https://github.com/openclaw/openclaw.git
# already have the repo? check out this version:$ git checkout v2026.2.19
└──▷ TRY IT
Remove a specific paired device from the gateway to revoke its access.
$ openclaw devices remove <device-id>
Clear all pending paired-device requests in one shot — useful during a security sweep or device reset.
$ openclaw devices clear --yes--pending
›Adds Apple Watch companion MVP with watch inbox UI, notification relay, and gateway command surfaces for watch status and send flows.
›Adds openclaw devices remove and openclaw devices clear --yes [--pending] CLI commands plus device.pair.remove gateway flow for paired-device hygiene.
›Adds APNs push registration and notification-signing configuration for iOS node delivery.
›Adds a push-test pipeline (push.test) for APNs delivery validation in gateway flows.
›Adds gateway.http.no_auth security audit findings when gateway.auth.mode="none" exposes Gateway HTTP APIs, with loopback-warning and remote-exposure-critical severities.
+1 moreshow less
›Enables wake of disconnected iOS nodes via APNs before nodes.invoke and auto-reconnects gateway sessions on silent push wake to reduce backgrounded-app invoke failures.
Semantic Kernel vector data adds SQL Server hybrid search, score threshold support, key auto-generation, and PostgreSQL DateTime/hybrid search improvements.
└──▷ GET THIS VERSION
$ git clone --branch vectordata-dotnet-10.0.0 https://github.com/microsoft/semantic-kernel.git
# already have the repo? check out this version:$ git checkout vectordata-dotnet-10.0.0
└──▷ USE IT
Restrict outbound HTTP calls from an SK Python agent to a known set of domains.
python
from semantic_kernel.core_plugins.http_plugin import HttpPlugin
http_plugin = HttpPlugin(allowed_domains=["api.example.com", "data.internal.corp"])
kernel.add_plugin(http_plugin, plugin_name="http")
›Implements ScoreThreshold support across MEVD providers, letting callers filter vector search results to only those meeting a minimum similarity score.
›Adds hybrid search to the SQL Server MEVD provider, combining vector and keyword search in a single query.
›Adds hybrid search to the PostgreSQL MEVD provider.
›Implements key auto-generation in MEVD, allowing records to be inserted without manually supplying a key.
Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.
Claude Code v2.1.49 adds file editing in Simple mode, ConfigChange hooks for auditing, and richer SDK model capability discovery.
└──▷ GET THIS VERSION
$ git clone --branch v2.1.49 https://github.com/anthropics/claude-code.git
# already have the repo? check out this version:$ git checkout v2.1.49
└──▷ TRY IT
Use Simple mode for lightweight file-editing tasks without needing a full agent setup.
$ CLAUDE_CODE_SIMPLE=1 claude
›Simple mode (CLAUDE_CODE_SIMPLE) now includes the file edit tool alongside the Bash tool, enabling direct file editing without leaving simple mode.
›New ConfigChange hook event fires when configuration files change mid-session, enabling enterprise security auditing and optional blocking of settings changes.
›SDK model info now exposes supportsEffort, supportedEffortLevels, and supportsAdaptiveThinking fields so consumers can programmatically discover model capabilities.
›Permission suggestions are now populated when safety checks trigger an ask response, enabling SDK consumers to surface permission options to users.
›Pressing Ctrl+C or ESC twice within 3 seconds now kills all background agents when the main thread is idle.
+1 moreshow less
›Improved permission prompts for path safety and working directory blocks now display the reason for the restriction.
└──▷ BREAKING ON UPGRADE
!Sonnet 4.5 with 1M context is being removed from the Max plan in favor of Sonnet 4.6 (which now has 1M context); users must switch via /model.
Pi v0.54.0 adds auto-discovery of agent skills from .agents/skills in project and home directories.
└──▷ GET THIS VERSION
$ git clone --branch v0.54.0 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:$ git checkout v0.54.0
›Adds automatic skill discovery from .agents/skills in the current directory, ancestor directories (up to git repo root), and ~/.agents/skills globally, alongside existing .pi skill paths.
›Adds tokenizer support for the LFM2.5-Audio-1.5B model architecture, convertible via python3 convert_hf_to_gguf.py /path/to/LFM2.5-Audio-1.5B/audio_detokenizer.
›Adds LFM2 SWA (Sliding Window Attention) model support alongside the new tokenizer.
llama.cpp b8100 adds full Modern BERT support including mean and rank pooling with GELU op.
└──▷ GET THIS VERSION
$ git clone --branch b8100 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:$ git checkout b8100
›Adds full Modern BERT model support, including mean pooling, rank pooling with GELU op, and a dense first layer as per the Modern BERT research paper, convertible via convert_hf_to_gguf.py.
llama.cpp b8099 adds FP16 MMA path on PowerPC for Q4/Q8 matmul, delivering 1.5–2x prompt-processing speedup.
└──▷ GET THIS VERSION
$ git clone --branch b8099 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:$ git checkout b8099
›Adds FP16×FP16→FP32 MMA acceleration path on PowerPC for Q4 and Q8 matrix multiplication, achieving 1.5–2x improvement in prompt-processing speed for Q4_0 and Q8_0 models on CPU.
$ git clone --branch arize-phoenix-v13.2.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:$ git checkout arize-phoenix-v13.2.0
›Adds AWS Bedrock cross-region inference model prefix preference, enabling selection of cross-region routing when calling Bedrock models.
Chroma 1.5.1 adds quantized SPANN indexing, FTS opt-out in schema, Advanced Search API goes GA, and new concurrency controls for log-service.
└──▷ GET THIS VERSION
$ git clone --branch 1.5.1 https://github.com/chroma-core/chroma.git
# already have the repo? check out this version:$ git checkout 1.5.1
›Enables users to disable full-text search (FTS) per collection via the schema — keeps index overhead away from collections that don't need text search.
›Adds a fetch_log semaphore configuration to control concurrency when fetching logs, preventing resource exhaustion under high compaction load.
›Adds a separate concurrency limit for manifest loads in log-service, independently capping the number of in-flight manifest operations.
›Promotes the Advanced Search API out of beta — now considered stable and no longer carries the beta label.
›Introduces a quantized SPANN segment writer and reader, wired into compaction and query orchestration — enables quantized approximate nearest-neighbor search for reduced memory and faster queries.
+5 moreshow less
›Adds garbage collection for usearch index files, reclaiming disk space from stale vector index segments.
›Adds delete_collection support for multi-collection multi-region (MCMR) deployments.
›Rejects fork_collection calls targeting multi-region databases with an explicit error rather than silently misbehaving.
›Makes the dirty_log_collections metric MCMR-aware for accurate observability in multi-region setups.
›Moves the compaction cursor into the WAL3 manifest as an intrinsic cursor, improving consistency of log compaction state.
THE AI TOOLCHAIN