菜单
Codex-native context and usage conservation mode. Use when the user asks for context conservative mode, limit conservation, cheap-model routing, or reducing context/token burn while keeping work moving.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | anti-hall-context-conserve |
| description | Codex-native context and usage conservation mode. Use when the user asks for context conservative mode, limit conservation, cheap-model routing, or reducing context/token burn while keeping work moving. |
This is the Codex port of anti-hall limit/context conservation. It combines hook nudges, model routing, and output hygiene; it does not depend on Claude Workflow JS.
Codex does not expand ${PLUGIN_ROOT} inside a skill's own instructions — that
variable is only set for plugin-bundled hook commands (see
docs/KB-codex-platform-hooks-plugins.md). Codex does show you this skill's own
file path when it selects the skill ("Codex starts with each skill's name,
description, and file path" — official Codex Skills doc). Resolve the plugin
root from that path before running anything below:
# SKILL_FILE = the absolute path Codex showed you for this SKILL.md.
ANTI_HALL_ROOT="$(cd "$(dirname "$SKILL_FILE")/../../.." && pwd)"
test -f "$ANTI_HALL_ROOT/.codex-plugin/plugin.json" || { echo "anti-hall plugin root not found relative to $SKILL_FILE — aborting" >&2; exit 1; }
Project-local hooks are installed by anti-hall-activate:
node "$ANTI_HALL_ROOT/codex/install-codex.js"
The installer registers hooks/limit-conserve-inject.js on UserPromptSubmit so conservation instructions are injected when active.
Manual controls:
export ANTIHALL_LIMIT_CONSERVE=on # force conservation
export ANTIHALL_LIMIT_CONSERVE=off # disable conservation
export ANTIHALL_LIMIT_THRESHOLD=85 # default threshold
Status check:
node -e "const {isConserving}=require('$ANTI_HALL_ROOT/hooks/limit-conserve.js'); console.log(JSON.stringify(isConserving(), null, 2))"
When conservation is active and the main/coordinator agent is on gpt-5.6-sol, switch it to gpt-5.6-terra (1M-context variant only) to preserve the flagship weekly bucket.
gpt-5.6-terra at 1M context — same context window, lower cost.gpt-5.4-mini (only 400k context; context loss is unacceptable).gpt-5.6-sol for delegated hard-reasoning subagents and on-demand escalation.gpt-5.6-terra and gpt-5.6-sol incur a confirmed 2× input / 1.5×
output multiplier once a request exceeds ~272K input tokens — this applies regardless of which
of the two you're on. For a very-large-context task (a big repo dump, a huge diff), consider
capping scope or routing to a Claude model instead (no equivalent premium up to 1M there). See
docs/KB-token-usage-models.md §2/§7.When conservation is active:
explore / gpt-5.4-mini for codebase lookup (gpt-5.6-luna when 5.6-era capability/cutoff matters; gpt-5.3-codex-spark is a distinct, faster/less-capable model, ChatGPT Pro only)gpt-5.4-mini (default) for mechanical command runninggpt-5.6-terra for settled implementationgpt-5.6-sol for planning, root-cause, validation, code review, and debate gatesgraphify-out/ exists.Codex does not expose Claude's Agent/Task PreToolUse hook surface for hard model-routing blocks. The Codex port enforces this via skill instructions, AGENTS.md, and UserPromptSubmit nudges instead of pretending it can block every spawn.
Explain anti-hall's optional DevSwarm integration from Codex — the hivecontrol reference KB, the (designed, unbuilt) workspace-tier orchestration, and the shipped layered recovery model (Claude-side companion: child self-report → supervisor poke → escalate-to-parent, automatic path NEVER kills, plus the on-demand devswarm-recover CLI — the only kill path). Use when the user asks about DevSwarm, hivecontrol, or the anti-hall liveness supervisor while working in Codex.
Explain and activate anti-hall's optional DevSwarm integration — the hivecontrol reference KB, the (designed, unbuilt) workspace-tier orchestration, and the shipped layered recovery model (child self-report → supervisor poke → escalate-to-parent, automatic path NEVER kills) plus the on-demand devswarm-recover CLI (the only path that ever kills). Use when the user asks "explain the anti-hall DevSwarm integration", "how do I activate the DevSwarm supervisor", "what DevSwarm addons does anti-hall have", "tune the liveness supervisor", "recover a stuck DevSwarm workspace", or anything about hivecontrol / DevSwarm workspaces from an anti-hall angle.
Check or update anti-hall from the local marketplace clone. Use when the user asks to update anti-hall, check whether anti-hall is current, or refresh the Codex port files.
Update anti-hall to the latest released version and show the changelog delta. Use when the user says "update anti-hall", "/anti-hall:update", "upgrade the plugin", "is anti-hall up to date", or "check for an anti-hall update". Fast-forward-pulls the marketplace clone, mirrors the new version into the plugin cache, prints what changed, then has the user reload in-session via /reload-plugins (rarely, a harness build may require a restart — the skill says so when relevant).
Brief the agent that installs or operates anti-hall (in Codex/OMX) on how the whole system works — a DERIVED, live enumeration of every installed hook, the shipped skills, the DevSwarm coordination substrate (mechanical triggers, store, CLI, auto-safe migration), and a docs/KB map. Generated from the real hooks.json and files on disk, never a hardcoded list, so it can't drift. Use when the user asks "brief me on anti-hall", "what's in this build", "how does the whole system work", or when onboarding an agent to install/run anti-hall. For "is it actually installed in Codex / do the guards fire", use anti-hall-doctor.
Brief the agent that installs or operates anti-hall on how the whole system works — a DERIVED, live enumeration of every installed hook (grouped by event, each with its own one-line purpose), the shipped skills, the DevSwarm coordination substrate (mechanical triggers, store, CLI, auto-safe migration), and a docs/KB map. Generated from the actual hooks.json and the files on disk, never a hardcoded list, so it can't drift. Use when the user asks "brief me on anti-hall", "what's in this build", "how does the whole system work", "give me the anti-hall system overview", "what hooks/skills/substrate ship here", or when onboarding an agent to install or run anti-hall. For "is it actually working / do the guards fire", use the `doctor` skill instead.