菜单
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.
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
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.
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.
Iterative TRIO-debate + fix-wave protocol for hardening non-trivial PRs and changes before merge. Spawns a parallel Reviewer + Auditor + Critic trio, dispatches fix waves, loops until convergence (zero NEW P0 or P1 blockers). Use before merging anything with cross-file risk, cross-PR coordination, security-sensitive changes (auth, signing, redaction, prompt-injection defense), schema migrations, production-data touches, shell scripts, CI/workflow YAML, cross-repo coordination, or LLM prompt work — anywhere a silent failure would be expensive. Born from a real project where 7 iteration rounds caught 30+ bugs that solo review would have missed.
| name | anti-hall-update |
| description | 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. |
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 $SKILL_FILE — aborting" >&2; exit 1; }
Use the existing pure-Node update helper:
node "$ANTI_HALL_ROOT/skills/update/scripts/update.js" --check
node "$ANTI_HALL_ROOT/skills/update/scripts/update.js"
For Codex, also re-run the Codex hook installer after a successful update:
node "$ANTI_HALL_ROOT/codex/install-codex.js"
or for global Codex hooks:
node "$ANTI_HALL_ROOT/codex/install-codex.js" --global
Do not force-pull, rebase, or delete plugin cache directories. If the update helper reports a dirty tree, diverged branch, offline state, or missing marketplace clone, surface that result and stop.
After a successful update, also run the capability scan to find what's missing on this machine vs what this build ships:
node "$ANTI_HALL_ROOT/scripts/capability-scan.js"
Read-only — it never installs anything. It reports each opt-in capability (companions under companion/install-*.js, statusline, pending state migrations) as {name, available, active, how}. Present a concise available-vs-active summary:
state-migrations at active: false — run node "$ANTI_HALL_ROOT/scripts/migrate-state.js" (idempotent, safe to re-run) to fold it, same as the Claude-side update flow.devswarm-detect's isDevswarmActive(process.env) — true only
when DEVSWARM_REPO_ID is set, i.e. the current session really is a DevSwarm
workspace (do NOT trigger on machine-level descriptor presence alone; the
session might be running outside DevSwarm). If inside a DevSwarm session, run
its how command (node companion/install-devswarm-supervisor.js)
regardless of the capability scan's active value. The installer is
idempotent (launchctl unload && load on macOS / systemd reload on Linux),
so this both first-installs when absent and refreshes an already-installed
supervisor so the next sweep runs this build's code. Report it plainly
("DevSwarm session detected — installed/refreshed the liveness supervisor to
<version>"). Fail-open: an install/refresh failure is reported, never fatal
to the update. If NOT inside a DevSwarm session, do not install — just
note it's available. (Safe to do unprompted: the supervisor's automatic
sweep never kills — killing is only the separate on-demand
devswarm-recover CLI. Defense in depth: the installed daemon is inert
without work — it only ever acts on descriptors under
~/.anti-hall/devswarm/workspaces/, so it no-ops when DevSwarm isn't
actually running.)isDevswarmActive(process.env) branch. When inside a DevSwarm
session, also run its how command
(node companion/install-devswarm-ingest.js) regardless of the scan's
active value. It is idempotent (launchctl unload && load on macOS /
systemctl --user daemon-reload + restart on Linux), so it first-installs
when absent and refreshes an already-installed daemon to this build's code.
Unlike the supervisor (a periodic sweep), the ingest daemon runs
continuously — the unit uses KeepAlive/Restart=always to re-exec it on
exit; hence a distinct label (com.anti-hall.devswarm-ingest) and log
(~/.anti-hall/devswarm-ingest.log). It is the single native consumer that
wraps hivecontrol workspace monitor and folds messages into the store;
nothing else auto-starts it, so a DevSwarm session without it silently ingests
nothing. Safe to install unprompted and idempotently: the daemon takes an
O_EXCL single-consumer lock, so only ONE instance ever runs even if installs
race. Report it plainly ("DevSwarm session detected — installed/refreshed
the ingest daemon to <version>"). Fail-open: an install/refresh failure is
reported, never fatal to the update. If NOT inside a DevSwarm session, do
not install — just note it's available.active: false (e.g. mcp-reaper — no
active-integration signal) — guide, don't auto-install: print its how
command and let the user decide.active: 'unknown' — the probe couldn't determine state (fail-open); mention it as unverified rather than claiming either state.active: true for everything — say so briefly; no gaps to report.After updating, restart Codex or start a fresh session if plugin/skill discovery does not reflect the new files immediately.