GitHub BenniAlencar/agent-cyber-sentinel (Apache-2.0): Tier-0 runtime defense for autonomous agents and MCP servers—prompt-injection/jailbreak firewall, secret leak interceptor with Shannon-entropy redaction, SSRF metadata blocking, tool-call sanitization, and tamper-evident Proof-of-Defense (POD-9) envelopes. Framed as the missing security layer when agents take real actions; CI + pytest suite in-repo.
Show HN (Oct 17): agentfork records every LLM and tool call, then lets you fork a failed run at any step and sweep prompts, models, or tool results with pass rates instead of anecdotes. Earlier steps replay from a local cache so you only pay for the step under test; OpenAI/Anthropic wrappers plus LangGraph support. Headline demo: a refund agent that reads cents as dollars keeps failing on bigger models until one system-prompt line fixes it (3/3 pass). pip install agentfork-cli; traces stay on-disk in .agentfork/traces.db.
Anthropic’s Applied AI playbook reframes Plan/Design/Build/Test/Deploy/Maintain as a loop of committed artifacts (intent.md → spec.md → plan.md → PR → production) with Claude Code plan mode, skills as institutional knowledge, hooks as deterministic gates, continuous evals in CI, and production control bands that write the next intent.md. Humans stay at judgment gates; agents do the rest. Companion pieces cover securing an AI-native SDLC and Claude Academy course modules.
junegunn (fzf author) documents how shell CTRL-T bindings die once you live in an agent prompt—and how tmux PREFIX bindings restore fzf: tmux-fzf-complete (files/dirs/URLs/on-screen paths/sentences) and fzf-git.sh (commits, branches, PRs) inject selections as keystrokes into Claude Code or Codex. Requires tmux 3.7+ floating panes and agent settings that keep scrollback (/tui default for Claude Code; Codex alternate_screen=never). Pattern: make agent prompts as completable as the shell.
Kondo’s WhatsApp MCP (Streamable HTTP at mcp.trykondo.com/mcp) lets Claude, ChatGPT, Cursor, Claude Code and Copilot search chats, read messages, draft replies, label, reminder and archive while Kondo Desktop holds the local WhatsApp session—no send tool, so nothing leaves as a message until you press Send. Relay keeps chat content off the cloud; Desktop must stay open. MCP on Business/Enterprise plans; install skill at trykondo.com/skills/kondo-whatsapp.md.
Managoat turns coding agents into a conversational computer API: POST a prompt, a sandbox wakes with repos/packages/credentials, Claude Code or Codex or Gemini CLI runs, then the machine parks (disk kept, no awake compute) until the next turn. Agent/Environment/Vault templates; secrets merge with Vault winning; bound secrets never enter the sandbox (egress broker). REST API, CLI, TypeScript SDK, demos (Conversations + Workbench), self-host docs. GitHub managoat/managoat actively shipping Oct 17.
GitHub BenniAlencar/agent-cyber-sentinel (Apache-2.0): Tier-0 runtime defense for autonomous agents and MCP servers—prompt-injection/jailbreak firewall, secret leak interceptor with Shannon-entropy redaction, SSRF metadata blocking, tool-call sanitization, and tamper-evident Proof-of-Defense (POD-9) envelopes. Framed as the missing security layer when agents take real actions; CI + pytest suite in-repo.
Show HN (Oct 17): agentfork records every LLM and tool call, then lets you fork a failed run at any step and sweep prompts, models, or tool results with pass rates instead of anecdotes. Earlier steps replay from a local cache so you only pay for the step under test; OpenAI/Anthropic wrappers plus LangGraph support. Headline demo: a refund agent that reads cents as dollars keeps failing on bigger models until one system-prompt line fixes it (3/3 pass). pip install agentfork-cli; traces stay on-disk in .agentfork/traces.db.
Anthropic’s Applied AI playbook reframes Plan/Design/Build/Test/Deploy/Maintain as a loop of committed artifacts (intent.md → spec.md → plan.md → PR → production) with Claude Code plan mode, skills as institutional knowledge, hooks as deterministic gates, continuous evals in CI, and production control bands that write the next intent.md. Humans stay at judgment gates; agents do the rest. Companion pieces cover securing an AI-native SDLC and Claude Academy course modules.
junegunn (fzf author) documents how shell CTRL-T bindings die once you live in an agent prompt—and how tmux PREFIX bindings restore fzf: tmux-fzf-complete (files/dirs/URLs/on-screen paths/sentences) and fzf-git.sh (commits, branches, PRs) inject selections as keystrokes into Claude Code or Codex. Requires tmux 3.7+ floating panes and agent settings that keep scrollback (/tui default for Claude Code; Codex alternate_screen=never). Pattern: make agent prompts as completable as the shell.