The AI Toolchain — issue -083, May 28, 2026

AutoGPT Platform v0.6.62 adds public agent chat sharing, Copilot bot commands, a Bots settings page, chat search, and cost breakdown visibility.

$ git clone --branch autogpt-platform-beta-v0.6.62 https://github.com/Significant-Gravitas/AutoGPT.git
# already have the repo? check out this version:
$ git checkout autogpt-platform-beta-v0.6.62

$ /resume

$ /new

• ›Adds CHAT_SEARCH feature flag to enable a chat search modal in the Copilot interface.
• ›Adds /resume command in the Copilot bot to pick and resume past chat sessions.
• ›Adds /new command in the Copilot bot to start a fresh conversation.
• ›Adds a Bots settings page with per-platform link management for connected bot integrations.
• ›Enables sharing agent chat results via a public link.

Framework for orchestrating role-playing, autonomous AI agents. By fostering collaborative intelligence, CrewAI empowers agents to work together seamlessly, tackling complex tasks.

CrewAI 1.14.6 adds env-var isolation in StdioTransport, Databricks env_var declarations, and enhanced planning configuration.

$ git clone --branch 1.14.6 https://github.com/crewAIInc/crewAI.git
# already have the repo? check out this version:
$ git checkout 1.14.6

• ›Adds env_vars declaration support on DatabricksQueryTool for explicit credential scoping.
• ›Improves planning configuration and observation handling in agent execution.
• ›Moves Skills Repository behind the CREWAI_EXPERIMENTAL gate as an experimental feature.

The agent that grows with you

Hermes v2026.5.28 adds multi-agent Kanban swarms, promptware defense, Bitwarden secret management, 4500× faster session search, and a curated MCP catalog.

$ git clone --branch v2026.5.28 https://github.com/NousResearch/hermes-agent.git
# already have the repo? check out this version:
$ git checkout v2026.5.28

$ hermes kanban swarm "Refactor the auth module and write tests"

$ hermes mcp

• ›Adds hermes kanban swarm command to auto-decompose a task into a full Swarm v1 graph (root, parallel workers, gated verifier, gated synthesizer, shared blackboard) with per-task model overrides, scheduled start times, and retry fingerprinting.
• ›Adds promptware/Brainworm-class injection defense: ~15 new threat patterns in tools/threat_patterns.py, memory scanning at load time, and delimiter markers on tool results to prevent impersonation of system content.
• ›Rebuilds session_search into a no-LLM, three-mode tool (discovery ~20ms, scroll ~1ms, browse) — eliminating ~$0.30/call and ~90s latency.
• ›Adds Bitwarden Secrets Manager integration: set BWS_ACCESS_TOKEN and Hermes pulls all provider credentials from Bitwarden at startup, with secrets.bitwarden.override_existing toggle and EU/self-hosted server support.
• ›Adds ntfy as the 23rd messaging platform plugin — no signup or API key required, just a topic URL, enabling push notifications from cron jobs, kanban completions, or send_message.

Build resilient agents.

LangGraph SDK 0.4.0 adds WebSocket streaming, reconnect resilience, scoped subgraph handles, and sync/async thread stream helpers.

$ git clone --branch sdk==0.4.0 https://github.com/langchain-ai/langgraph.git
# already have the repo? check out this version:
$ git checkout sdk==0.4.0

• ›Adds WebSocket stream transport support, enabling lower-latency bidirectional streaming as an alternative to SSE.
• ›Adds SSE transport and v3 streaming primitives as a foundational streaming layer.
• ›Adds WebSocket stream selection wiring so clients can choose between SSE and WebSocket transports.
• ›Adds async stream reconnect support with hardened reconnect logic for resilient long-running streams.
• ›Adds async and sync thread stream helpers for high-level, ergonomic consumption of streamed thread output.

Your own personal AI assistant. Any OS.

OpenClaw v2026.5.27 adds native OpenAI-compatible embeddings, Pixverse video generation, full DeepInfra catalog, and plugin approval metadata.

$ git clone --branch v2026.5.27 https://github.com/openclaw/openclaw.git
# already have the repo? check out this version:
$ git checkout v2026.5.27

• ›Adds a core OpenAI-compatible embedding provider supporting local and hosted OpenAI-style endpoints, with config, doctor, and docs integration.
• ›Adds the Pixverse video generation provider with API region selection, docs, and external plugin packaging support.
• ›Loads the full credential-aware DeepInfra model catalog when users browse models during onboarding, preserving configured API-key catalogs.
• ›Exposes plugin approval action metadata via the Plugin SDK.
• ›Adds plugin display metadata to ClawHub so catalog and package listings use cleaner names.

Semantic Kernel dotnet-1.77.0 enables server URL validation for OpenAPI plugins and ships Agent Framework 1.0-compatible migration samples.

$ git clone --branch dotnet-1.77.0 https://github.com/microsoft/semantic-kernel.git
# already have the repo? check out this version:
$ git checkout dotnet-1.77.0

• ›Enables default-on server URL validation for OpenAPI plugins, adding a security guardrail that checks server URLs before requests are made.
• ›Publishes Agent Framework 1.0-compatible migration samples and updates, aligning Semantic Kernel Agent Framework usage with the 1.0 release.

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 v2.1.154 adds Opus 4.8 with fast mode, dynamic multi-agent workflows, shell background sessions, and smarter plugin discovery.

$ git clone --branch v2.1.154 https://github.com/anthropics/claude-code.git
# already have the repo? check out this version:
$ git checkout v2.1.154

$ claude
# Inside the session:
> Create a workflow to audit every Go package in this repo for security issues
> /workflows

$ claude agents
# Inside agents view:
! npm run build:watch

$ claude --bg --exec 'pytest tests/ -x -q'

• ›Adds Claude Opus 4.8 as the new default model with /effort xhigh for maximum effort on hard tasks.
• ›Introduces dynamic workflows: ask Claude to orchestrate work across tens to hundreds of background agents; view runs with /workflows.
• ›Fast mode on Opus 4.8 now available at 2× the standard rate for 2.5× the speed.
• ›The lean system prompt is now the default for all models except Haiku, Sonnet, and Opus 4.7 and earlier.
• ›Reduces unsolicited multiple-choice prompts: Claude now reserves them for decisions it genuinely cannot make itself.

$ git clone --branch v2.1.153 https://github.com/anthropics/claude-code.git
# already have the repo? check out this version:
$ git checkout v2.1.153

# In your Claude Code marketplace config:
github:
  repo: owner/heavy-dataset-repo
  skipLfs: true

$ # status_line.sh — now receives COLUMNS and LINES automatically
printf '%*s' "$COLUMNS" '' | tr ' ' '-'

Autonomous coding agent as an SDK, IDE extension, or CLI assistant.

Cline v3.86.0 adds Claude Opus 4.8 (with 1M-context variants) and Moonshot Kimi K2 model support.

$ git clone --branch v3.86.0 https://github.com/cline/cline.git
# already have the repo? check out this version:
$ git checkout v3.86.0

• ›Adds Claude Opus 4.8 provider support, including 1M-context variants where available.

Glamourous agentic coding for all

Crush v0.74.0 adds SSH notifications, diff view for denied tools, and deeper client-server session sharing

$ git clone --branch v0.74.0 https://github.com/charmbracelet/crush.git
# already have the repo? check out this version:
$ git checkout v0.74.0

$ CRUSH_CLIENT_SERVER=1 crush

• ›Adds diff view for denied tools so practitioners can inspect exactly what a tool was attempting before deciding to allow or block it.
• ›Supports notifications over SSH terminals, enabling alerts when running Crush on remote hosts.
• ›Adds configurable notification backend with bell support and a backend picker UI.
• ›Exposes an in-progress flag and connected-client count on session API responses, enabling richer monitoring of active sessions.
• ›Server now shares one workspace per directory across all connected clients, reducing duplication in multi-client setups.

GitHub Copilot CLI v1.0.55 adds /autopilot, per-MCP token visibility, Claude Opus 4.8, and new permissions controls.

$ git clone --branch v1.0.55 https://github.com/github/copilot-cli.git
# already have the repo? check out this version:
$ git checkout v1.0.55

permissions:
  disableBypassPermissionsMode: true

$ /autopilot Refactor the authentication module to use OAuth 2.0 PKCE flow

session.create({ pluginDirectories: ["/path/to/my-plugins"] })

• ›Adds permissions.disableBypassPermissionsMode setting to prevent users from enabling allow-all/yolo mode.
• ›Adds /autopilot <objective> command (with /goal as an alias) to keep autopilot sessions focused on a declared objective.
• ›Adds pluginDirectories parameter to session.create and session.resume RPC calls, letting SDK clients mount Open Plugins-format directories per session.
• ›Shows per-MCP-server token usage in /mcp and breaks out MCP tool tokens separately in /context.
• ›Adds --plugin-dir skill precedence ordering: project > plugin-dir > personal > custom, so --plugin-dir skills now override personal-home (~/.copilot, ~/.agents) skills of the same name.

Lightweight coding agent that runs in your terminal

Codex CLI gains richer doctor diagnostics, Vim text-object editing, named permission profiles, and non-interactive install support.

$ git clone --branch rust-v0.135.0 https://github.com/openai/codex.git
# already have the repo? check out this version:
$ git checkout rust-v0.135.0

$ codex doctor

$ CODEX_NON_INTERACTIVE=1 bash install.sh

$ /status

• ›Adds richer environment, Git, terminal, app-server, and thread inventory diagnostics to codex doctor for support cases.
• ›New /status command shows remote connection details and server version when connected over a remote transport.
• ›Vim mode gains text-object editing, improved word/line-end behavior, and a configurable turn-interruption keybind.
• ›Adds named permission profile picker to /permissions, which now displays configured custom profiles.
• ›Packaged Codex builds can discover and use the bundled patched zsh helper on supported macOS and Linux targets.

The open source coding agent.

OpenCode v1.15.12 adds WebSocket transport for OpenAI, ACP prompt/slash-command support, workspace management dialog, and a tab layout setting.

$ git clone --branch v1.15.12 https://github.com/sst/opencode.git
# already have the repo? check out this version:
$ git checkout v1.15.12

$ OPENCODE_EXPERIMENTAL_WEBSOCKETS=true opencode

• ›Adds WebSocket transport for OpenAI responses, enabled via OPENCODE_EXPERIMENTAL_WEBSOCKETS=true.
• ›Enables ACP integrations to send prompts, slash commands, and usage updates through acp-next.
• ›Adds a workspace management dialog to the TUI.
• ›Adds a Desktop setting to switch tab layouts.

AI agent toolkit: unified LLM API, agent loop, TUI, coding agent CLI

Pi v0.77.0 adds selective tool disablement, headless Codex device-code login, streaming-aware extension input, and Claude Opus 4.8 support.

$ git clone --branch v0.77.0 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:
$ git checkout v0.77.0

$ pi --exclude-tools bash

• ›Adds --exclude-tools / -xt flag to disable specific built-in, extension, or custom tools while keeping all others active.
• ›Adds device-code auth for headless ChatGPT Plus/Pro Codex subscription login via /login.
• ›Adds InputEvent.streamingBehavior so extensions can distinguish idle prompts, mid-stream steers, and queued follow-ups.
• ›Adds Claude Opus 4.8 model metadata and updated adaptive-thinking coverage for Anthropic.

llama.cpp b9383 adds the Granite 4.1 chat template for IBM Granite model support.

$ git clone --branch b9383 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b9383

• ›Adds Granite 4.1 chat template, enabling correct prompt formatting for IBM Granite 4.1 models.

$ git clone --branch b9380 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b9380

LanceDB python-v0.33.0 adds namespace management, LSM write spec, unenforced primary keys, and streaming ingestion primitives.

$ git clone --branch python-v0.33.0 https://github.com/lancedb/lancedb.git
# already have the repo? check out this version:
$ git checkout python-v0.33.0

• ›Adds order_by method to the Node.js Query class for sorting query results.
• ›Adds progress callback to Table.add in Node.js to track ingestion progress.
• ›Adds renameTable method on Node.js Connection for in-place table renaming.
• ›Adds namespace management methods on Node.js Connection for creating, listing, and deleting namespaces.
• ›Exposes connectNamespace on Node.js Connection for namespace-backed connections.

$ git clone --branch v0.30.0 https://github.com/lancedb/lancedb.git
# already have the repo? check out this version:
$ git checkout v0.30.0

await table.add(data, { progress: (count) => console.log(`Inserted ${count} rows`) });

Langfuse v3.176.0 adds in-app agent API keys, full-text search, code-based evals, MCP expansions, and dashboard import/export.

$ git clone --branch v3.176.0 https://github.com/langfuse/langfuse.git
# already have the repo? check out this version:
$ git checkout v3.176.0

• ›Adds import/export for dashboard widget configs, enabling sharing and reuse of dashboard layouts.
• ›Exposes metrics, scores, media, comments, datasets, annotation queues, and a health route via MCP — broadening the MCP surface for agent integrations.
• ›Adds code-based eval evaluator type, allowing custom evaluation logic written in code; supports manual batch runs on inactive evaluators.
• ›Adds in-app agent API key management, letting agents authenticate without leaving the Langfuse UI.
• ›Adds initial full-text search (FTS) implementation with string filters that always use the TEXT index.

Weave v0.52.41 links Gen AI OTel spans to Evals, adds agent custom attribute schema queries, and auto-writes 'latest' aliases on object creation.

$ git clone --branch v0.52.41 https://github.com/wandb/weave.git
# already have the repo? check out this version:
$ git checkout v0.52.41

• ›Adds WEAVE_INSECURE_DISABLE_SSL environment variable support to the OTLP Exporter for deployments requiring SSL bypass.
• ›Links Gen AI spans to Evals automatically via a new OTel span processor, enabling eval association without manual wiring.
• ›Adds agent custom attribute schema queries, allowing structured querying of agent-specific attributes.
• ›Writes an explicit latest alias on every obj_create call, so the most recent object version is always addressable by alias.
• ›Infers JSON filter casts in the feedback query path, improving filter expressiveness for feedback data.