codex

codex is a command for Claude Code from pwdev-solucoes/pwdev-claude-marketplace. It costs 16 tokens per session (902 once invoked), scanned A, original, Apache-2.0.

A command that sends a coding task to the Codex command-line tool and then checks the changes it produces. It can either allow edits or request a read-only analysis.

In plain words
What is it for?
Use it for clearly scoped code changes, bug fixes, or tests. Use read-only mode when you want analysis without allowing files to be changed.
Why use it?
It gives you a repeatable way to delegate a focused implementation or bug fix while still requiring a review of the result. It also records the starting state so changes can be checked against it.

Command for Claude Code

Written for Claude Code: allowed-tools in frontmatter. Also seen: mentions Codex.

Runs only inside its plugin — its command needs a path that Claude Code sets for a plugin’s own hooks and for nothing else. Install the plugin, not this.

Part of the pwdev-code plugin — 2 skills, 22 commands, 8 agents, 6 hooks shipped together

Install

Getting it into your agent

This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.

Claude Code
/plugin marketplace add pwdev-solucoes/pwdev-claude-marketplace
Claude Code
/plugin install pwdev-code

Made for: Claude Code.

Or install pwdev-code, the plugin that ships this one along with the rest of its 2 skills, 22 commands, 8 agents, 6 hooks.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for codex

README.md
[![agentmods](https://agentmods.dev/badge/commands/pwdev-solucoes/pwdev-claude-marketplace/codex.svg)](https://agentmods.dev/commands/pwdev-solucoes/pwdev-claude-marketplace/codex)
Your own site
<a href="https://agentmods.dev/commands/pwdev-solucoes/pwdev-claude-marketplace/codex"><img src="https://agentmods.dev/badge/commands/pwdev-solucoes/pwdev-claude-marketplace/codex.svg" alt="Measured on agentmods" height="20"></a>
Per session 16 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 902 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 5.1 $0.00016 $0.00902
Opus 5 $0.00008 $0.00451
Sonnet 5 $0.00003 $0.00180
Haiku 4.5 $0.00002 $0.00090

Measured 6d ago against content hash 86ede6044a97, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

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 6d 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/pwdev-code/commands/codex.md · 82 lines

How it starts

The opening of the file, as written. The whole thing — 82 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/pwdev-code:codex — Delegate to Codex CLI

Role

Orchestrator: sends the task to the external Codex CLI via the standardized delegation script, then performs a MANDATORY critical review of the result. Best for: objective implementation, bugfixes, well-scoped tests. Protocol details: ${CLAUDE_PLUGIN_ROOT}/references/delegation.md

Input

$ARGUMENTS: the task description. --read-only → MODE=read (default: write).

Flow

STEP 0 — Language

Follow ${CLAUDE_PLUGIN_ROOT}/references/language.md (resolve lang from .planning/config.json; ask only if unset).

STEP 1 — Preflight

  • Parse $ARGUMENTS: strip --read-only (→ MODE=read); the remainder is the task. Empty task → show usage and STOP.
  • git status --porcelain — if the tree is dirty in write mode, warn the user that delegation results will be hard to attribute to the agent; offer to continue or abort.
  • Record baseline: git rev-parse HEAD and git status --short.

STEP 2 — Human Confirmation (same rule as review.md STEP 3.5)

Show the EXACT command that will run:

"${CLAUDE_PLUGIN_ROOT}/scripts/run-agent.sh" codex <write|read> "<task>"

If this is the FIRST external CLI run in this session → require explicit confirmation before executing. Never proceed silently.

STEP 3 — Run

Execute the command via Bash with a tool timeout ≥ the script timeout + 60s (default script timeout: 600s; override via external_models.codex.timeout_s).

  • exit 127 → Codex CLI not installed (npm i -g @openai/codex, then authenticate once by running codex); STOP.
  • exit 124 → report the timeout; suggest raising external_models.codex.timeout_s in .planning/config.json.
  • exit 4 → another write delegation is running; wait for it — NEVER run two write delegations at once.
  • Output copy (when .planning/ exists): .planning/delegation/<ts>-codex.md

STEP 4 — MANDATORY Review Protocol (never skip, even on exit 0)

  1. git status --short — list every touched file.
  2. git diff --stat, then read the FULL git diff (per-file if large).
  3. Scope check: flag any file outside the task scope; flag ANY touch of .env*, secrets, lockfiles not implied by the task, or CI configs.
  4. Run the project's relevant test suite yourself — do not trust the agent's claim.
  5. Critical evaluation with YOUR OWN judgment: correctness, conventions, simplicity, missing edge cases — do not rubber-stamp the agent's summary.
  6. NEVER commit or push. If the result is bad, PROPOSE (do not execute without confirmation) git checkout -- <files> / git clean -n.

Read the full file on GitHub · 82 lines

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. 6d ago First seen · 82 lines · 16 tokens per session scan A 86ede6044a97

Subscribe to this mod's changes

codex is a command published in the GitHub repository pwdev-solucoes/pwdev-claude-marketplace (2 stars, last pushed 17d ago), licensed Apache-2.0. It adds 16 tokens to every session and 902 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.