<?xml version="1.0" encoding="UTF-8"?>
<feed xmlns="http://www.w3.org/2005/Atom">
  <title>terminalblog</title>
  <subtitle>Following the AI coding agent ecosystem — what ships, what's debated, and what matters. Claude Code, Cursor, Hermes, Codex, and everything in between.</subtitle>
  <link href="https://terminalblog.com/rss.xml" rel="self" />
  <link href="https://terminalblog.com/" />
  <updated>2026-07-12T10:00:00.000Z</updated>
  <id>https://terminalblog.com/</id>
  <author>
    <name>Anshad</name>
  </author>
  <icon>https://terminalblog.com/favicon.svg</icon>
  
    <entry>
      <id>https://terminalblog.com/blog/gpt-5-6-claude-fable-code-arena/</id>
      <title><![CDATA[GPT-5.6 Sol Matches Claude Fable 5 on Code Arena — For 40% Less]]></title>
      <link href="https://terminalblog.com/blog/gpt-5-6-claude-fable-code-arena/" />
      <published>2026-07-12T10:00:00.000Z</published>
      <updated>2026-07-12T10:00:00.000Z</updated>
      <summary type="html"><![CDATA[GPT-5.6 Sol ties Claude Fable 5 on the Code Arena benchmark at 40% lower cost, shaking up the performance-per-dollar calculation for coding agents.]]></summary>
      <author><name>Anshad</name></author>
      <category term="news" />
      <category term="benchmarks" />
      <category term="gpt" />
      <category term="claude" />
      <category term="coding-ai" />
      <category term="pricing" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/kraken-agentic-trading-bots/</id>
      <title><![CDATA[Kraken's Mobile Relaunch Puts Agentic Trading Bots in Your Pocket]]></title>
      <link href="https://terminalblog.com/blog/kraken-agentic-trading-bots/" />
      <published>2026-07-12T09:45:00.000Z</published>
      <updated>2026-07-12T09:45:00.000Z</updated>
      <summary type="html"><![CDATA[Kraken relaunches its mobile app with built-in agentic trading bots, letting users deploy automated strategies from their phones — no separate infrastructure needed.]]></summary>
      <author><name>Anshad</name></author>
      <category term="news" />
      <category term="launches" />
      <category term="agentic" />
      <category term="trading" />
      <category term="kraken" />
      <category term="crypto" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/ethereum-foundation-ai-audit-bugs/</id>
      <title><![CDATA[Ethereum Foundation Found Real Bugs With AI Audits — This Changes Smart Contract Security]]></title>
      <link href="https://terminalblog.com/blog/ethereum-foundation-ai-audit-bugs/" />
      <published>2026-07-12T09:30:00.000Z</published>
      <updated>2026-07-12T09:30:00.000Z</updated>
      <summary type="html"><![CDATA[The Ethereum Foundation used AI-powered audits to uncover real vulnerabilities in smart contract code — validating that coding agents can catch bugs that traditional review misses.]]></summary>
      <author><name>Anshad</name></author>
      <category term="news" />
      <category term="security" />
      <category term="ai-audits" />
      <category term="ethereum" />
      <category term="smart-contracts" />
      <category term="bugs" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/tencent-hy3-14-cents-coding-ai/</id>
      <title><![CDATA[Tencent Just Dropped a Coding AI at $0.14/M Tokens — Cheaper Than Your Coffee]]></title>
      <link href="https://terminalblog.com/blog/tencent-hy3-14-cents-coding-ai/" />
      <published>2026-07-12T09:15:00.000Z</published>
      <updated>2026-07-12T09:15:00.000Z</updated>
      <summary type="html"><![CDATA[Tencent's Hy3 coding AI launches at $0.14 per million input tokens, undercutting most frontier models and resetting expectations for per-token pricing in the coding agent space.]]></summary>
      <author><name>Anshad</name></author>
      <category term="news" />
      <category term="pricing" />
      <category term="coding-ai" />
      <category term="tencent" />
      <category term="benchmarks" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/bytedance-seedream-5-0-pro-platforms/</id>
      <title><![CDATA[ByteDance Just Flipped the Switch on Seedream 5.0 Pro — Multiplatform Release Is Live]]></title>
      <link href="https://terminalblog.com/blog/bytedance-seedream-5-0-pro-platforms/" />
      <published>2026-07-12T09:00:00.000Z</published>
      <updated>2026-07-12T09:00:00.000Z</updated>
      <summary type="html"><![CDATA[ByteDance rolls out Seedream 5.0 Pro across multiple platforms, bringing its latest image generation model to a wider audience — and giving coding agents another visual generation option.]]></summary>
      <author><name>Anshad</name></author>
      <category term="news" />
      <category term="launches" />
      <category term="image-generation" />
      <category term="ai-models" />
      <category term="seedream" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/claude-code-steganography/</id>
      <title><![CDATA[Claude Code's Steganographic Date Stamp — When Developer Tools Play Spy Games]]></title>
      <link href="https://terminalblog.com/blog/claude-code-steganography/" />
      <published>2026-07-12T00:00:00.000Z</published>
      <updated>2026-07-12T00:00:00.000Z</updated>
      <summary type="html"><![CDATA[A reverse engineer found Claude Code silently encodes API gateway info into system prompt punctuation. We unpack the article, the HN debate, and what it means for trust in AI coding agents.]]></summary>
      <author><name>Anshad</name></author>
      <category term="claude-code" />
      <category term="security" />
      <category term="privacy" />
      <category term="opinion" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/claude-code-win11-tui-frame-pacing/</id>
      <title><![CDATA[Claude Code Is Crashing Your Frame Rate on Windows 11 — The Fix]]></title>
      <link href="https://terminalblog.com/blog/claude-code-win11-tui-frame-pacing/" />
      <published>2026-07-12T00:00:00.000Z</published>
      <updated>2026-07-12T00:00:00.000Z</updated>
      <summary type="html"><![CDATA[A bisected regression shows Claude Code's TUI render loop dropped from 16 fps to ~9-10 fps on Windows 11 between versions 2.1.159 and 2.1.207. Here's how to check and fix the stutter.]]></summary>
      <author><name>Anshad</name></author>
      <category term="bug" />
      <category term="performance" />
      <category term="claude-code" />
      <category term="windows" />
      <category term="tui" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/claude-code-windows-click-through-permissions/</id>
      <title><![CDATA[Your Claude Code May Be Silently Approving Permissions — Here's How to Check]]></title>
      <link href="https://terminalblog.com/blog/claude-code-windows-click-through-permissions/" />
      <published>2026-07-12T00:00:00.000Z</published>
      <updated>2026-07-12T00:00:00.000Z</updated>
      <summary type="html"><![CDATA[A Windows click-to-focus bug in Claude Code causes the first click on a de-focused window to activate a pending permission dialog, submitting an unintended answer without the user knowing.]]></summary>
      <author><name>Anshad</name></author>
      <category term="security" />
      <category term="bug" />
      <category term="claude-code" />
      <category term="windows" />
      <category term="permissions" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/claude-code-steganography-trust-betrayal/</id>
      <title><![CDATA[Claude Code Has Been Embedding Steganographic Markers in Your Prompts — Here's the Full Story]]></title>
      <link href="https://terminalblog.com/blog/claude-code-steganography-trust-betrayal/" />
      <published>2026-07-11T22:00:00.000Z</published>
      <updated>2026-07-11T22:00:00.000Z</updated>
      <summary type="html"><![CDATA[Anthropic's Claude Code silently marks system prompts with invisible Unicode steganography to detect Chinese AI labs and API resellers. The HN community had 2,444 things to say.]]></summary>
      <author><name>Anshad</name></author>
      <category term="security" />
      <category term="claude-code" />
      <category term="privacy" />
      <category term="steganography" />
      <category term="opinion" />
      <category term="featured" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/claude-code-background-subagent-secret-leak/</id>
      <title><![CDATA[Your Background Subagents Can Leak Secrets — Build the Isolation Model]]></title>
      <link href="https://terminalblog.com/blog/claude-code-background-subagent-secret-leak/" />
      <published>2026-07-11T21:10:00.000Z</published>
      <updated>2026-07-11T21:10:00.000Z</updated>
      <summary type="html"><![CDATA[A freshly filed, reproducible Claude Code security issue shows background subagents stalling and emitting authorization-shaped prompt fragments. Here's the mental model for keeping secrets out of subagent output.]]></summary>
      <author><name>Anshad</name></author>
      <category term="security" />
      <category term="subagents" />
      <category term="claude-code" />
      <category term="isolation" />
      <category term="guide" />
      <category term="mental-model" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/tencent-hy3-coding-ai-cheap-input-tokens/</id>
      <title><![CDATA[Tencent's Hy3 Coding AI Puts Input Tokens at $0.14 Per Million]]></title>
      <link href="https://terminalblog.com/blog/tencent-hy3-coding-ai-cheap-input-tokens/" />
      <published>2026-07-11T21:00:00.000Z</published>
      <updated>2026-07-11T21:00:00.000Z</updated>
      <summary type="html"><![CDATA[A new coding-focused AI model from Tencent launched with input tokens priced at $0.14 per million — a price point that lands well below most frontier models and targets developers directly.]]></summary>
      <author><name>Anshad</name></author>
      <category term="news" />
      <category term="roundup" />
      <category term="pricing" />
      <category term="coding-ai" />
      <category term="tencent" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/hermes-just-shipped-gateway-runtime-context-budgets/</id>
      <title><![CDATA[Hermes Just Shipped Gateway Runtime Context Budget Fixes — @ References Now Scope Correctly]]></title>
      <link href="https://terminalblog.com/blog/hermes-just-shipped-gateway-runtime-context-budgets/" />
      <published>2026-07-11T18:35:00.000Z</published>
      <updated>2026-07-11T18:35:00.000Z</updated>
      <summary type="html"><![CDATA[10 commits ship across Hermes Agent fixing gateway runtime context budgets, @ reference scope expansion, compaction anti-thrash stale verdicts, and Codex manual compaction error handling.]]></summary>
      <author><name>Anshad</name></author>
      <category term="hermes" />
      <category term="just-shipped" />
      <category term="release" />
      <category term="fixes" />
      <category term="gateway" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/openclaw-just-shipped-fleet-cell-supervisor/</id>
      <title><![CDATA[OpenClaw Just Shipped Fleet Cell Supervisor — Multi-Tenant Hosting Arrives]]></title>
      <link href="https://terminalblog.com/blog/openclaw-just-shipped-fleet-cell-supervisor/" />
      <published>2026-07-11T18:35:00.000Z</published>
      <updated>2026-07-11T18:35:00.000Z</updated>
      <summary type="html"><![CDATA[OpenClaw's fleet cell supervisor delivers multi-tenant hosting for agent instances, plus browser profile administration, Feishu streaming timeouts, and macOS cookie-import improvements — 10 commits landed.]]></summary>
      <author><name>Anshad</name></author>
      <category term="openclaw" />
      <category term="just-shipped" />
      <category term="release" />
      <category term="features" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/hermes-just-shipped-gateway-runtime-readiness/</id>
      <title><![CDATA[Hermes Just Shipped Gateway Parent Runtime Session Scope]]></title>
      <link href="https://terminalblog.com/blog/hermes-just-shipped-gateway-runtime-readiness/" />
      <published>2026-07-11T18:00:00.000Z</published>
      <updated>2026-07-11T18:00:00.000Z</updated>
      <summary type="html"><![CDATA[Hermes Agent lands authenticated runtime readiness checks, session-scoped model picker fixes, and gateway startup notifications — 4 commits tightening reliability for multi-session setups.]]></summary>
      <author><name>Anshad</name></author>
      <category term="hermes" />
      <category term="just-shipped" />
      <category term="release" />
      <category term="gateway" />
      <category term="fixes" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/oh-my-pi-just-shipped-custom-role-management/</id>
      <title><![CDATA[oh-my-pi Just Shipped Custom Role Management in Model Hub]]></title>
      <link href="https://terminalblog.com/blog/oh-my-pi-just-shipped-custom-role-management/" />
      <published>2026-07-11T18:00:00.000Z</published>
      <updated>2026-07-11T18:00:00.000Z</updated>
      <summary type="html"><![CDATA[oh-my-pi lands custom role management and a tiered search session selector — 2 commits that make model configuration and session recall faster.]]></summary>
      <author><name>Anshad</name></author>
      <category term="oh-my-pi" />
      <category term="just-shipped" />
      <category term="release" />
      <category term="features" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/openclaw-just-shipped-claude-session-fleet/</id>
      <title><![CDATA[OpenClaw Just Shipped Claude Session Fleet — Here's What It Does]]></title>
      <link href="https://terminalblog.com/blog/openclaw-just-shipped-claude-session-fleet/" />
      <published>2026-07-11T18:00:00.000Z</published>
      <updated>2026-07-11T18:00:00.000Z</updated>
      <summary type="html"><![CDATA[OpenClaw's biggest push this week brings a Claude session fleet, pinned SSH tunnel runtime for cloud workers, Slack data viz, and workspace-directory plugins for Codex — 10 commits, 4 major features.]]></summary>
      <author><name>Anshad</name></author>
      <category term="openclaw" />
      <category term="just-shipped" />
      <category term="release" />
      <category term="features" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/genlayer-internet-court-ai-agents-27-firms/</id>
      <title><![CDATA[27 Firms Just Backed the World's First Internet Court for AI Agents]]></title>
      <link href="https://terminalblog.com/blog/genlayer-internet-court-ai-agents-27-firms/" />
      <published>2026-07-11T17:30:00.000Z</published>
      <updated>2026-07-11T17:30:00.000Z</updated>
      <summary type="html"><![CDATA[GenLayer launches an Internet Court for AI agents, backed by 27 firms — a governance layer for resolving disputes between autonomous agents operating on-chain.]]></summary>
      <author><name>Anshad</name></author>
      <category term="news" />
      <category term="genlayer" />
      <category term="governance" />
      <category term="ai-agents" />
      <category term="industry" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/claude-code-permission-prompt-flood/</id>
      <title><![CDATA[Your Claude Code May Ask 700+ Permission Prompts Per Session — Here's How to Check]]></title>
      <link href="https://terminalblog.com/blog/claude-code-permission-prompt-flood/" />
      <published>2026-07-11T17:25:00.000Z</published>
      <updated>2026-07-11T17:25:00.000Z</updated>
      <summary type="html"><![CDATA[Claude Code's compound-command permission system can flood you with hundreds of prompts per session, even for read-only commands like cd, ls, and git status. Here's the workaround.]]></summary>
      <author><name>Anshad</name></author>
      <category term="bug" />
      <category term="claude-code" />
      <category term="security" />
      <category term="permissions" />
      <category term="windows" />
      <category term="featured" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/cursor-sand-ai-office-agent-challenge-anthropic/</id>
      <title><![CDATA[Cursor Is Building an Office Agent That Does More Than Code — and Anthropic Is in Its Sights]]></title>
      <link href="https://terminalblog.com/blog/cursor-sand-ai-office-agent-challenge-anthropic/" />
      <published>2026-07-11T17:25:00.000Z</published>
      <updated>2026-07-11T17:25:00.000Z</updated>
      <summary type="html"><![CDATA[Cursor's Sand AI office agent expands beyond code editing into the full developer workflow — email, docs, tickets — directly challenging Anthropic's agent turf.]]></summary>
      <author><name>Anshad</name></author>
      <category term="news" />
      <category term="cursor" />
      <category term="office-agent" />
      <category term="anthropic" />
      <category term="industry" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/gpt-5-6-sol-ultra-64-subagents-prove-math-conjecture/</id>
      <title><![CDATA[GPT-5.6 Sol Ultra Just Used 64 Subagents to Crack a 50-Year-Old Math Problem]]></title>
      <link href="https://terminalblog.com/blog/gpt-5-6-sol-ultra-64-subagents-prove-math-conjecture/" />
      <published>2026-07-11T17:20:00.000Z</published>
      <updated>2026-07-11T17:20:00.000Z</updated>
      <summary type="html"><![CDATA[OpenAI's GPT-5.6 Sol Ultra deployed 64 coordinated subagents to prove a long-standing math conjecture — a milestone in multi-agent reasoning at scale.]]></summary>
      <author><name>Anshad</name></author>
      <category term="news" />
      <category term="research" />
      <category term="gpt-5-6" />
      <category term="math" />
      <category term="subagents" />
      <category term="openai" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/oh-my-pi-crashing-macos-native-grep/</id>
      <title><![CDATA[Oh My Pi Is Crashing macOS With Native Grep — The Fix]]></title>
      <link href="https://terminalblog.com/blog/oh-my-pi-crashing-macos-native-grep/" />
      <published>2026-07-11T17:20:00.000Z</published>
      <updated>2026-07-11T17:20:00.000Z</updated>
      <summary type="html"><![CDATA[Oh My Pi's native grep can crash the entire agent process when a file changes during search. Here's how to check if you're affected and what to do about it.]]></summary>
      <author><name>Anshad</name></author>
      <category term="bug" />
      <category term="crash" />
      <category term="oh-my-pi" />
      <category term="macos" />
      <category term="grep" />
      <category term="featured" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/ai-news-roundup-2026-07-11/</id>
      <title><![CDATA[AI News Roundup: Grok 4.5 Hits Tesla, Perplexity's Orchestrator Beats Opus, and Meta Undercuts Pricing]]></title>
      <link href="https://terminalblog.com/blog/ai-news-roundup-2026-07-11/" />
      <published>2026-07-11T17:15:00.000Z</published>
      <updated>2026-07-11T17:15:00.000Z</updated>
      <summary type="html"><![CDATA[Today's roundup: Musk pushes Grok 4.5 inside Tesla and SpaceX, Perplexity's orchestrator tops Opus on a benchmark, Meta launches a cheap API, ByteDance ships Seedream 5.0 Pro, and Cursor builds an office agent.]]></summary>
      <author><name>Anshad</name></author>
      <category term="news" />
      <category term="industry" />
      <category term="benchmarks" />
      <category term="launches" />
      <category term="roundup" />
      <category term="featured" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/unsloth-qwen3-6-quantizations-2-5x-gpu-speed/</id>
      <title><![CDATA[Unsloth's New Qwen3.6 Quantizations Run 2.5x Faster on Your Existing GPU]]></title>
      <link href="https://terminalblog.com/blog/unsloth-qwen3-6-quantizations-2-5x-gpu-speed/" />
      <published>2026-07-11T17:15:00.000Z</published>
      <updated>2026-07-11T17:15:00.000Z</updated>
      <summary type="html"><![CDATA[Unsloth ships Qwen3.6 quantizations delivering 2.5x GPU speed — a performance jump that lets teams run larger models on the same hardware without upgrading.]]></summary>
      <author><name>Anshad</name></author>
      <category term="news" />
      <category term="launches" />
      <category term="quantization" />
      <category term="gpu" />
      <category term="performance" />
      <category term="unsloth" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/gitlawb-zero-0-4-0-npm-sandbox/</id>
      <title><![CDATA[gitlawb-zero Heads to 0.4.0 With a Native npm Binary and Tighter Sandbox]]></title>
      <link href="https://terminalblog.com/blog/gitlawb-zero-0-4-0-npm-sandbox/" />
      <published>2026-07-11T17:10:00.000Z</published>
      <updated>2026-07-11T17:10:00.000Z</updated>
      <summary type="html"><![CDATA[gitlawb-zero's 0.4.0 release prep ships its native binary as platform optionalDependencies, reworks Windows sandbox denial classification, and adds a GitHub Copilot provider.]]></summary>
      <author><name>Anshad</name></author>
      <category term="release" />
      <category term="gitlawb-zero" />
      <category term="sandbox" />
      <category term="npm" />
      <category term="devsecops" />
      <category term="featured" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/meta-muse-spark-1-1-api-25-percent-pricing/</id>
      <title><![CDATA[Meta Just Slashed AI API Pricing to a Quarter of What Everyone Else Charges]]></title>
      <link href="https://terminalblog.com/blog/meta-muse-spark-1-1-api-25-percent-pricing/" />
      <published>2026-07-11T17:10:00.000Z</published>
      <updated>2026-07-11T17:10:00.000Z</updated>
      <summary type="html"><![CDATA[Meta launches Muse Spark 1.1 API at roughly 25% of competitor pricing — a floor that changes the economics of routing for cost-sensitive coding agents.]]></summary>
      <author><name>Anshad</name></author>
      <category term="news" />
      <category term="pricing" />
      <category term="meta" />
      <category term="api" />
      <category term="launches" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/codex-gpt-5-6-tool-call-bugs/</id>
      <title><![CDATA[Codex Users Hit Tool-Call and Workspace Bugs on the New GPT-5.6 Build]]></title>
      <link href="https://terminalblog.com/blog/codex-gpt-5-6-tool-call-bugs/" />
      <published>2026-07-11T17:05:00.000Z</published>
      <updated>2026-07-11T17:05:00.000Z</updated>
      <summary type="html"><![CDATA[Fresh Codex issues report a namespace collision producing 'unsupported custom tool call: execexec' on gpt-5.6-sol, plus new tasks that start without workspace tools and a missing model on cloud workstations.]]></summary>
      <author><name>Anshad</name></author>
      <category term="bug" />
      <category term="codex" />
      <category term="gpt-5-6" />
      <category term="tool-calls" />
      <category term="windows" />
      <category term="featured" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/codex-sandbox-memory-consolidation/</id>
      <title><![CDATA[Codex Tightens Sandbox Enforcement for Memory Consolidation]]></title>
      <link href="https://terminalblog.com/blog/codex-sandbox-memory-consolidation/" />
      <published>2026-07-11T17:00:00.000Z</published>
      <updated>2026-07-11T17:00:00.000Z</updated>
      <summary type="html"><![CDATA[A merged Codex commit preserves parent sandbox enforcement during memory consolidation — closing a path where a sub-process could escape the boundaries the parent session set.]]></summary>
      <author><name>Anshad</name></author>
      <category term="bug" />
      <category term="security" />
      <category term="codex" />
      <category term="sandbox" />
      <category term="featured" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/openclaw-macos-launchd-crash-loop/</id>
      <title><![CDATA[OpenClaw's macOS Gateway Can Crash-Loop on Config Change]]></title>
      <link href="https://terminalblog.com/blog/openclaw-macos-launchd-crash-loop/" />
      <published>2026-07-11T16:55:00.000Z</published>
      <updated>2026-07-11T16:55:00.000Z</updated>
      <summary type="html"><![CDATA[A P0 issue reports that OpenClaw's macOS launchd gateway exits without relaunching on a config change, leaving the agent dead until a manual kickstart — flagged as a release-blocker.]]></summary>
      <author><name>Anshad</name></author>
      <category term="bug" />
      <category term="openclaw" />
      <category term="macos" />
      <category term="crash-loop" />
      <category term="gateway" />
      <category term="featured" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/openclaw-claude-fleet-cloud-workers/</id>
      <title><![CDATA[OpenClaw Adds a Claude Session Fleet and Production Cloud Workers]]></title>
      <link href="https://terminalblog.com/blog/openclaw-claude-fleet-cloud-workers/" />
      <published>2026-07-11T16:50:00.000Z</published>
      <updated>2026-07-11T16:50:00.000Z</updated>
      <summary type="html"><![CDATA[OpenClaw's latest commits introduce a Claude session fleet and production cloud-worker bundles with pinned SSH bootstrap and an admission handshake — a move toward running agents as infrastructure.]]></summary>
      <author><name>Anshad</name></author>
      <category term="release" />
      <category term="openclaw" />
      <category term="cloud" />
      <category term="fleet" />
      <category term="self-host" />
      <category term="featured" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/oh-my-pi-grok-build-provider-cpu-spin/</id>
      <title><![CDATA[oh-my-pi Adds a Grok Build Provider — and Users Report CPU Spin and RPC Crashes]]></title>
      <link href="https://terminalblog.com/blog/oh-my-pi-grok-build-provider-cpu-spin/" />
      <published>2026-07-11T16:45:00.000Z</published>
      <updated>2026-07-11T16:45:00.000Z</updated>
      <summary type="html"><![CDATA[Alongside a new isolated Grok Build subscription provider, oh-my-pi's issue tracker shows a ~50% idle CPU spin, an RPC mode that crashes on bad stdin, and a Bun binary hardcoded to CI paths.]]></summary>
      <author><name>Anshad</name></author>
      <category term="bug" />
      <category term="oh-my-pi" />
      <category term="cpu" />
      <category term="crash" />
      <category term="providers" />
      <category term="featured" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/oh-my-pi-model-hub-session-selector/</id>
      <title><![CDATA[oh-my-pi Ships a Model Hub and a Faster Session Selector]]></title>
      <link href="https://terminalblog.com/blog/oh-my-pi-model-hub-session-selector/" />
      <published>2026-07-11T16:40:00.000Z</published>
      <updated>2026-07-11T16:40:00.000Z</updated>
      <summary type="html"><![CDATA[A burst of commits to oh-my-pi's coding agent adds a unified model hub with custom roles, spatial sidebar navigation, a tiered session selector, and tighter mid-prompt skill matching.]]></summary>
      <author><name>Anshad</name></author>
      <category term="release" />
      <category term="oh-my-pi" />
      <category term="model-hub" />
      <category term="tui" />
      <category term="featured" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/hermes-secret-leakage-sandbox-windows-failures/</id>
      <title><![CDATA[Two Hermes Bugs Worth Watching: Secret Leakage in Redaction and Silent Windows Failures]]></title>
      <link href="https://terminalblog.com/blog/hermes-secret-leakage-sandbox-windows-failures/" />
      <published>2026-07-11T16:35:00.000Z</published>
      <updated>2026-07-11T16:35:00.000Z</updated>
      <summary type="html"><![CDATA[Fresh issue reports flag a secret-redaction leak in worktree handling and Windows command failures being misclassified as sandbox denials — both on the same upstream agent.]]></summary>
      <author><name>Anshad</name></author>
      <category term="bug" />
      <category term="security" />
      <category term="hermes" />
      <category term="data-leak" />
      <category term="windows" />
      <category term="featured" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/hermes-gateway-parent-runtime-session-scope/</id>
      <title><![CDATA[Hermes Hardens Its Gateway: Live Runtime Checks and Session-Scoped Model Switches]]></title>
      <link href="https://terminalblog.com/blog/hermes-gateway-parent-runtime-session-scope/" />
      <published>2026-07-11T16:30:00.000Z</published>
      <updated>2026-07-11T16:30:00.000Z</updated>
      <summary type="html"><![CDATA[A sweep of recent Hermes Agent commits tightens how the gateway decides a session is ready and makes mid-session model switches stick to the right scope — plus better usage attribution.]]></summary>
      <author><name>Anshad</name></author>
      <category term="release" />
      <category term="hermes" />
      <category term="gateway" />
      <category term="session-management" />
      <category term="featured" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/claude-code-rc-norton-360-silent-kill/</id>
      <title><![CDATA[Your Coding Agent Is Lying About Its Health — Here's How to Catch It]]></title>
      <link href="https://terminalblog.com/blog/claude-code-rc-norton-360-silent-kill/" />
      <published>2026-07-11T10:45:00.000Z</published>
      <updated>2026-07-11T10:45:00.000Z</updated>
      <summary type="html"><![CDATA[A Claude Code Remote Control bug revealed a deeper problem: your agent's status indicators can be confidently wrong. Here's a diagnostic framework that works across every coding agent on every platform.]]></summary>
      <author><name>Anshad</name></author>
      <category term="guide" />
      <category term="diagnostics" />
      <category term="claude-code" />
      <category term="windows" />
      <category term="networking" />
      <category term="remote-control" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/claude-code-sonnet-5-deletes-folders/</id>
      <title><![CDATA[Claude Code's Sonnet 5 Wiped an Entire Folder While Just Trying to List Its Files]]></title>
      <link href="https://terminalblog.com/blog/claude-code-sonnet-5-deletes-folders/" />
      <published>2026-07-11T07:30:00.000Z</published>
      <updated>2026-07-11T07:30:00.000Z</updated>
      <summary type="html"><![CDATA[A new bug report shows Anthropic's Sonnet 5 model, running inside Claude Code, deleting the entire contents of a folder as it tried to enumerate files — a silent data-loss event no prompt asked for.]]></summary>
      <author><name>Anshad</name></author>
      <category term="opinion" />
      <category term="trend" />
      <category term="claude-code" />
      <category term="bug" />
      <category term="data-loss" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/codex-crash-leaks-system-instructions/</id>
      <title><![CDATA[Codex App Crashes and Leaks Its Own System Instructions in the Error Message]]></title>
      <link href="https://terminalblog.com/blog/codex-crash-leaks-system-instructions/" />
      <published>2026-07-11T05:00:00.000Z</published>
      <updated>2026-07-11T05:00:00.000Z</updated>
      <summary type="html"><![CDATA[A reproducible crash in Codex Desktop spills internal system prompts into the error output — revealing exactly how the agent is instructed to behave.]]></summary>
      <author><name>Anshad</name></author>
      <category term="opinion" />
      <category term="trend" />
      <category term="codex" />
      <category term="security" />
      <category term="bug" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/windows-stepchild-coding-agents-fixing-up/</id>
      <title><![CDATA[Windows Is the Unloved Stepchild of Coding Agents — and That's Changing]]></title>
      <link href="https://terminalblog.com/blog/windows-stepchild-coding-agents-fixing-up/" />
      <published>2026-07-11T04:00:00.000Z</published>
      <updated>2026-07-11T04:00:00.000Z</updated>
      <summary type="html"><![CDATA[A burst of Windows-specific crash and stability fixes just landed across Hermes, Codex, and Goose at the same time. It's the most honest signal yet that the coding-agent category is growing up.]]></summary>
      <author><name>Anshad</name></author>
      <category term="opinion" />
      <category term="trend" />
      <category term="windows" />
      <category term="codex" />
      <category term="hermes" />
      <category term="goose" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/coding-agent-features-comparison-2026/</id>
      <title><![CDATA[Coding Agent Features in 2026: Complete Comparison Matrix]]></title>
      <link href="https://terminalblog.com/blog/coding-agent-features-comparison-2026/" />
      <published>2026-07-11T02:00:00.000Z</published>
      <updated>2026-07-11T02:00:00.000Z</updated>
      <summary type="html"><![CDATA[Vision, cron, multi-provider, git integration, plugins, subagents — how all 15 agents compare across every feature category.]]></summary>
      <author><name>Anshad</name></author>
      <category term="pillar" />
      <category term="guide" />
      <category term="features" />
      <category term="comparison" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/coding-agent-pricing-guide-2026/</id>
      <title><![CDATA[Coding Agent Pricing in 2026: What Each Agent Actually Costs]]></title>
      <link href="https://terminalblog.com/blog/coding-agent-pricing-guide-2026/" />
      <published>2026-07-11T01:00:00.000Z</published>
      <updated>2026-07-11T01:00:00.000Z</updated>
      <summary type="html"><![CDATA[Free vs subscription vs pay-as-you-go — the complete pricing breakdown for all 15 coding agents on terminalblog.]]></summary>
      <author><name>Anshad</name></author>
      <category term="pillar" />
      <category term="guide" />
      <category term="pricing" />
      <category term="comparison" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/open-source-vs-commercial-coding-agents-guide/</id>
      <title><![CDATA[Open Source vs Commercial Coding Agents: The Complete Guide]]></title>
      <link href="https://terminalblog.com/blog/open-source-vs-commercial-coding-agents-guide/" />
      <published>2026-07-11T00:00:00.000Z</published>
      <updated>2026-07-11T00:00:00.000Z</updated>
      <summary type="html"><![CDATA[A detailed comparison of every open-source and commercial coding agent tracked on terminalblog — pricing, licensing, community, and long-term viability.]]></summary>
      <author><name>Anshad</name></author>
      <category term="pillar" />
      <category term="guide" />
      <category term="comparison" />
      <category term="deep-dive" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/multi-agent-orchestration-over-bigger-models/</id>
      <title><![CDATA[Why I'm Betting on Multi-Agent Orchestration Over Bigger Models]]></title>
      <link href="https://terminalblog.com/blog/multi-agent-orchestration-over-bigger-models/" />
      <published>2026-07-10T23:30:00.000Z</published>
      <updated>2026-07-10T23:30:00.000Z</updated>
      <summary type="html"><![CDATA[A single smarter model is the obvious path. Coordinating multiple smaller models is the path that actually ships more code.]]></summary>
      <author><name>Anshad</name></author>
      <category term="opinion" />
      <category term="orchestration" />
      <category term="multi-agent" />
      <category term="architecture" />
      <category term="prediction" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/coding-agent-fleet-not-single-tool/</id>
      <title><![CDATA[Your Next Coding Agent Will Be a Fleet, Not a Single Tool]]></title>
      <link href="https://terminalblog.com/blog/coding-agent-fleet-not-single-tool/" />
      <published>2026-07-10T23:00:00.000Z</published>
      <updated>2026-07-10T23:00:00.000Z</updated>
      <summary type="html"><![CDATA[The era of one agent to rule them all is ending. The future is 5-10 specialized agents working together, each good at one thing.]]></summary>
      <author><name>Anshad</name></author>
      <category term="opinion" />
      <category term="prediction" />
      <category term="multi-agent" />
      <category term="fleet" />
      <category term="ecosystem" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/feature-every-coding-agent-missing/</id>
      <title><![CDATA[The Feature Every Coding Agent Is Missing]]></title>
      <link href="https://terminalblog.com/blog/feature-every-coding-agent-missing/" />
      <published>2026-07-10T22:00:00.000Z</published>
      <updated>2026-07-10T22:00:00.000Z</updated>
      <summary type="html"><![CDATA[Every agent can write code. None of them can tell you what your codebase actually needs. That gap is the biggest opportunity in AI coding tools.]]></summary>
      <author><name>Anshad</name></author>
      <category term="opinion" />
      <category term="feature-gap" />
      <category term="future" />
      <category term="codebase-analysis" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/automate-code-review-ai-agents/</id>
      <title><![CDATA[How I Automate My Entire Code Review Process With AI Agents]]></title>
      <link href="https://terminalblog.com/blog/automate-code-review-ai-agents/" />
      <published>2026-07-10T21:00:00.000Z</published>
      <updated>2026-07-10T21:00:00.000Z</updated>
      <summary type="html"><![CDATA[Code review is the bottleneck that agents are perfectly suited to solve. Here's a concrete workflow that runs every day without human intervention.]]></summary>
      <author><name>Anshad</name></author>
      <category term="guide" />
      <category term="code-review" />
      <category term="automation" />
      <category term="workflow" />
      <category term="cron-jobs" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/copilot-cli-terminal-agent-github-repos/</id>
      <title><![CDATA[GitHub Copilot CLI: The Terminal Agent That Knows Your Repos]]></title>
      <link href="https://terminalblog.com/blog/copilot-cli-terminal-agent-github-repos/" />
      <published>2026-07-10T20:00:00.000Z</published>
      <updated>2026-07-10T20:00:00.000Z</updated>
      <summary type="html"><![CDATA[GitHub's terminal agent brings deep repository integration to the command line — and it's cheaper than you think.]]></summary>
      <author><name>Anshad</name></author>
      <category term="opinion" />
      <category term="copilot-cli" />
      <category term="github" />
      <category term="terminal" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/ampcode-vs-copilot-cli-pricing-battle/</id>
      <title><![CDATA[AmpCode vs Copilot CLI: Pay-As-You-Go vs Subscription]]></title>
      <link href="https://terminalblog.com/blog/ampcode-vs-copilot-cli-pricing-battle/" />
      <published>2026-07-10T19:00:00.000Z</published>
      <updated>2026-07-10T19:00:00.000Z</updated>
      <summary type="html"><![CDATA[Two terminal agents with fundamentally different pricing models — which one makes financial sense?]]></summary>
      <author><name>Anshad</name></author>
      <category term="comparison" />
      <category term="ampcode" />
      <category term="copilot-cli" />
      <category term="pricing" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/unspoken-problem-ai-coding-tools/</id>
      <title><![CDATA[The Unspoken Problem With AI Coding Tools Right Now]]></title>
      <link href="https://terminalblog.com/blog/unspoken-problem-ai-coding-tools/" />
      <published>2026-07-10T19:00:00.000Z</published>
      <updated>2026-07-10T19:00:00.000Z</updated>
      <summary type="html"><![CDATA[Every new tool promises 10x productivity. What nobody mentions is the context debt, the tool sprawl, and the growing dependency on models you don't control.]]></summary>
      <author><name>Anshad</name></author>
      <category term="opinion" />
      <category term="ecosystem" />
      <category term="problems" />
      <category term="dependency" />
      <category term="viral" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/coding-agents-security-hardening-wave/</id>
      <title><![CDATA[Coding Agents Just Got Serious About Security — and It's About Time]]></title>
      <link href="https://terminalblog.com/blog/coding-agents-security-hardening-wave/" />
      <published>2026-07-10T18:05:00.000Z</published>
      <updated>2026-07-10T18:05:00.000Z</updated>
      <summary type="html"><![CDATA[A wave of prompt-injection and data-leak hardening just landed across OpenClaw, Goose, Hermes, and Codex. The coding agent is now an attack surface, and the ecosystem is finally treating it like one.]]></summary>
      <author><name>Anshad</name></author>
      <category term="opinion" />
      <category term="trend" />
      <category term="openclaw" />
      <category term="goose" />
      <category term="hermes" />
      <category term="codex" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/agent-not-ready-shared-infrastructure/</id>
      <title><![CDATA[Why Your Coding Agent Isn't Ready to Be Shared Infrastructure]]></title>
      <link href="https://terminalblog.com/blog/agent-not-ready-shared-infrastructure/" />
      <published>2026-07-10T18:00:00.000Z</published>
      <updated>2026-07-10T18:00:00.000Z</updated>
      <summary type="html"><![CDATA[We're quietly shipping coding agents onto servers, into gateways, and behind CI — but the trust model underneath them is still built for one person on one laptop.]]></summary>
      <author><name>Anshad</name></author>
      <category term="opinion" />
    </entry>
  

    <entry>
      <id>https://terminalblog.com/blog/best-coding-agent-setup-six-months/</id>
      <title><![CDATA[The Best Coding Agent Setup I've Found After 6 Months]]></title>
      <link href="https://terminalblog.com/blog/best-coding-agent-setup-six-months/" />
      <published>2026-07-10T18:00:00.000Z</published>
      <updated>2026-07-10T18:00:00.000Z</updated>
      <summary type="html"><![CDATA[Not one agent. Not two. A layered system that handles everything from quick edits to complex refactors to scheduled maintenance.]]></summary>
      <author><name>Anshad</name></author>
      <category term="opinion" />
      <category term="workflow" />
      <category term="setup" />
      <category term="layered" />
      <category term="guides" />
    </entry>
  
</feed>