Ecosystem Update - 2026-05-28

TL;DR

• One safe Quick Win was implemented: restored allow_login_shell = false in the opt-in conservative profiles after the setting had drifted out of ~/.codex/config.toml.
• Official Codex rust-v0.134.0 is available while local CLI is 0.133.0; the release adds local conversation-history search, profile migration guidance, read-only MCP concurrency, and richer hook context.
• Today's source crawl reinforces the current posture: keep native memories disabled until explicitly piloted, avoid wholesale community plugin imports, and evolve hooks only from stable Codex payloads.

Quick Wins

Auto-Implemented

• Backed up config.toml, hooks.json, and current agent TOML files under /Users/chadsimon/.codex/backups/2026-05-28/.
• Updated /Users/chadsimon/.codex/config.toml so both opt-in conservative profiles disable login-shell semantics.
• Verified config.toml with Python tomllib.
• Verified hooks.json with python3 -m json.tool /Users/chadsimon/.codex/hooks.json.
• Verified posture with python3 /Users/chadsimon/.codex/bin/codex_config_posture.py --mode warn; it reported Codex config posture ok.
• Ran python3 /Users/chadsimon/.codex/bin/codex-runtime-doctor --summary --ascii; result was errors=0 warnings=1, with the only warning being 4 stale/temp trusted project entries.
• Ran TERM=xterm-256color codex --strict-config doctor --summary --ascii; result was 13 ok | 1 idle | 3 notes | 0 warn | 0 fail ok.
• A plain codex --strict-config doctor --summary --ascii was also attempted and failed only because this noninteractive shell exposes TERM=dumb.

Build Queue

• Stable Codex 0.134.0 upgrade and smoke (Codex-md) - https://github.com/openai/codex/releases/tag/rust-v0.134.0 - Local Doctor reports 0.134.0 available; upgrade should be a focused runtime change with post-upgrade hooks, MCP, plugin, and doctor smoke checks.
• Native conversation-history search migration decision (Codex-md) - https://github.com/openai/codex/releases/tag/rust-v0.134.0 - Codex now has official local conversation-history search; evaluate whether python3 ~/.codex/bin/codex-session-search should remain canonical, become a wrapper, or be kept for JSON/transcript-specific recall.
• Hook subagent identity intake (hook) - https://github.com/openai/codex/releases/tag/rust-v0.134.0 - New hook inputs include subagent identity. Add a tiny fixture/test pass before using it for per-subagent verification ledgers or failure context.
• Memory root boundary audit (mcp) - https://github.com/openai/codex/commit/d5ec93f - Upstream moved native memories root setup out of core config. Current setup keeps features.memories = false and uses omni-mem; audit only if a native-memory pilot is reopened.
• Auto-review profile value normalization (Codex-md) - https://developers.openai.com/codex/config-reference - Docs list approvals_reviewer = "auto_review"; local conservative-auto-review still uses the legacy guardian_subagent alias. Normalize later with a compatibility check, not as a drive-by edit.
• CTF architecture eval adapter (research) - https://arxiv.org/abs/2605.21497 - The paper shows structured specialist roles can improve consistency/cost in security tasks; map it to codex-security and security-audit evals before adding new agents.

Research

• Autonomous LLM Agents & CTFs: A Second Look - Directly relevant to security-agent evaluation because it compares general-purpose agents with engineered specialist architectures across 30 web CTF tasks.

Already Have

Power-user default model/profile posture, prompt telemetry off, live web search, official config schema header, native hooks enabled, UserPromptSubmit route classification, Bash PreToolUse safety guard, Bash PostToolUse verification and failure-context hooks, SessionStart cached repo context and config-posture checks, Stop and PreCompact omni-mem hooks, OpenAI developer docs MCP, omni-mem MCP, Browser/Chrome/Computer Use/Documents/Spreadsheets/Presentations/Gmail/OpenAI Developers plugins, read-only explorer/planner/reviewer/python-reviewer/typescript-reviewer/validator agents, scoped workspace-write worker and chad-twin agents, bounded agent depth/thread/runtime caps, conservative/review profiles, conservative login-shell hardening restored, destructive app tools disabled globally, pokegen disabled through [[skills.config]], session recall, skill audit, planning-gate, /auto, build/drive/go/govern wrappers, security-audit, codex-security, runtime doctor, what-would-chad-do, daily ecosystem state tracking, prior self-evolving-agent and agent-worm research intake, prior hook parity watch items, and prior community skill-catalog audit items.

Rejected

• Auto-upgrade Codex to 0.134.0 as a Quick Win - rejected: upgrading the active runtime is a larger state change than a harness hardening edit and needs a focused smoke window.
• Enable native Codex memories - rejected: current memory authority is omni-mem, and today's upstream memory-root boundary change is a reason to keep native memory pilots explicit.
• Enable plugin_hooks globally - rejected: hook contribution from plugins is still a trust boundary and remains intentionally off.
• Wholesale import from Awesome Claude Code, Claude Code Toolkit, or Codex community skill catalogs - rejected: outside skills/plugins require strict audit and the local library already covers the recurring workflows.
• Global auto-format PostToolUse hook - rejected: formatting is repo-specific and should not be enforced globally from a community pattern.
• Install agent/session manager stacks such as ORCH, ccmanager, obey, Bouncer, or pro-workflow - rejected: current AgentOps, hooks, runtime doctor, and reviewer agents cover the recurring local needs without adding another orchestration layer.
• Turn Codex Python SDK beta commits into local runtime work - rejected: today's upstream Python SDK commits are release-packaging/documentation changes, not local harness gaps.

Sources checked: https://github.com/hesreallyhim/awesome-claude-code, https://howborisusesclaudecode.com/, https://github.com/shanraisshan/codex-cli-best-practice, https://github.com/rohitg00/awesome-claude-code-toolkit, https://developers.openai.com/codex/config-reference, https://developers.openai.com/codex/cli/slash-commands#built-in-slash-commands, https://github.com/openai/codex/releases, https://github.com/openai/codex/commit/d5ec93f, https://arxiv.org/search/?searchtype=all&query=LLM+agent+coding&order=-announced_date_first, https://arxiv.org/abs/2605.21497, https://arxiv.org/abs/2604.27264, https://arxiv.org/abs/2605.02812, web search: "Codex new hooks agents skills site:github.com 2026", web search: "arxiv.org LLM agent coding autonomous 2026 site:arxiv.org" Tier 2 fetched: yes Tier 3 fetched: yes Run at: 2026-05-28T10:35:04Z