watch-codex

A command for reading text aloud through CVI, a Claude voice-integration tool, using its configured voice and speech settings.

In plain words
What is it for?
Use it to speak a supplied piece of text, optionally choosing a voice or reading speed.
Why use it?
It lets you hear selected text without manually copying it into another text-to-speech tool.

Command

Part of the utils plugin — 2 commands shipped together

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/signalcompose/claude-tools/watch-codex
Clone the repo
git clone --depth 1 https://github.com/signalcompose/claude-tools

Or install utils, the plugin that ships this one along with the rest of its 2 commands.

Per session 13 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 952 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00013 $0.00952
Opus 5 $0.00006 $0.00476
Sonnet 5 $0.00003 $0.00190
Haiku 4.5 $0.00001 $0.00095

Measured 3d ago against content hash 3bdc61d1fd57, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

watch-codex scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

plugins/utils/commands/watch-codex.md · 71 lines

What it actually says

Watch Codex Job

Codex companion のジョブを監視し、進捗・停滞・外部 kill を行として出力する。

なぜ必要か

codex-companion.mjs status --jsonstatus を待機条件にしてはいけない。 ジョブが外部から kill されても "status": "running" のまま残る(dead pid の回収処理が 存在しない)ため、until status != running のループは永久に発火しない。 2026-08-27 にこれで 64 分を失った。

このコマンドは companion を呼ばず state.json を直接読み、ログの mtime と pid の生存 という2つの事実で判定する。

Usage

引数なしで、現在の workspace の実行中ジョブを監視する:

!bash ${CLAUDE_PLUGIN_ROOT}/scripts/watch-codex.sh --help

watch-codex.sh [job-id] [--stall-secs N] [--interval N] [--workspace PATH] [--state-root PATH]

job-idcodex-companion.mjs status --jsonrunning[].id、または state.jsonjobs[].id発注直後に監視を張るなら job-id を明示するのが確実 (省略時は「その workspace の最新 active ジョブ」を選ぶため、起動が早すぎると まだ対象が存在しない)。

出力する行と exit code

意味 exit
MILESTONE コマンド完了/失敗(exit code つき) -
PHASE editing -> verifying などの遷移 -
STALL ログが N 秒伸びていない・プロセスは生存 2
KILLED pid 消滅・state は running のまま 3
DONE status=completed で正常終了 0
FAILED status=failed/cancelled 等で終了 4
ERROR 監視対象を解決できない / 監視を継続できない 1

成功と失敗を同じ exit 0 に畳まない。畳むと呼び出し側が終了コードで区別できず、 「running でなくなった=成功」という元の誤りに戻る。

起動時点で完了済みのマイルストーンは再生しない(件数と直近1件だけ表示)。

回し方

Monitor tool に食わせるか、run_in_background で回す。 出力行がそのまま通知したい事象になっている。

🔴 STALL が出た時の心得

「N 分応答なし」を即ハングと断定しない。 STALLKILLED を分けているのは、 原因が違えば対処も違うため。

事象 意味 対処
KILLED プロセスが消滅している ハングではない。再開する。原因調査に時間を使わない
STALL プロセスは生きているがログが伸びない 長時間コマンドかハング。まず当のコマンドを自分で回して切り分ける

2026-08-27 の実例: 停滞していたコマンドを手元で実行したら exit 0・数秒で完了した。 つまりハングではなく外部 kill だった。切り分けを飛ばして「ハングの原因調査」に 入っていたら、さらに時間を失っていた。

詳細は ${CLAUDE_PLUGIN_ROOT}/references/codex-stall-triage.md を読む。

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago First seen · 71 lines · 13 tokens per session scan A 3bdc61d1fd57

Subscribe to this mod's changes

watch-codex is a command published in the GitHub repository signalcompose/claude-tools (4 stars, last pushed 6d ago), licensed MIT. It adds 13 tokens to every session and 952 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.