Name: Codex Bridge
Author: xallindrome-labs

Codex Bridge MCP

Lets Claude Code consult the Codex CLI for plans and images, as durable background jobs.

Grade A Local · stdio No auth needed 5 tools v0.1.0 MIT

No reviews yet by xallindrome-labsnpm Runs on your machine, nothing hosted

93/100

trust score

tools declared

local

runs on your machine

—

downloads a day

What it does

Lets Claude Code consult the Codex CLI for plans and images, as durable background jobs.

Quickstart

# 1 — add it to Claude Code
claude mcp add --scope user xallindrome-labs-codex-bridge-mcp '--' npx -y @xallindrome-labs/codex-bridge

# 2 — ask your agent for something these tools do
#   codex_plan, codex_image, codex_result and 2 more

Codex Bridge costs nothing on mcprush: there is no plan to choose here, no cap to set and nothing mcprush can bill you for.

Publishernot claimed

xallindrome-labs1 server

1 category

All 1 server

Collected from a public index. Nobody has claimed this account, so nothing here was written by its author — claim it if it is yours.

Installation

Where are you running it?

Every route below installs the same thing and ends at the same approval screen. This one runs on your machine: your client starts Codex Bridge as a process under your own user, with your files and your network, so the tool surface below is what it can do to your computer rather than to a server somewhere else. Its scan result is on the Security tab — read the surface before you approve it.

Claude Codedeclared by this listing

This is a public server: you run it yourself and this marketplace is not in the path. You need Node.js — the current LTS (24) is the safe choice; npx comes with it. Claude Code registers it in one command; --scope user makes it available in every project. The '--' is in quotes so that PowerShell passes it on.

claude mcp add --scope user xallindrome-labs-codex-bridge-mcp '--' npx -y @xallindrome-labs/codex-bridge

Reconnect, or start a new session, and the tools appear in the model’s tool list.

Free on mcprush, and it runs from its publisher’s own package: this marketplace is not in the path, so there is no key of ours to hold and no card to add.

What it asks of you

Nothing is required, but it can take CODEX_BIN (Path to the Codex CLI binary. Auto-detected when unset.), CODEX_BRIDGE_HOME (Where job files are stored. Default: ~/.codex-bridge), CODEX_BRIDGE_IDLE_TIMEOUT_MS (Stop a run after this many ms with no Codex output. Default: 600000), CODEX_BRIDGE_PLAN_MODEL (Model for codex_plan. Default: your Codex config.), CODEX_BRIDGE_PLAN_EFFORT (Reasoning effort for codex_plan. Default: your Codex config.), CODEX_BRIDGE_IMAGE_MODEL (Model for codex_image. Default: gpt-6-luna), CODEX_BRIDGE_IMAGE_EFFORT (Reasoning effort for codex_image. Default: high) if you want to set them.

What it writes

One config entry your client uses to start the process locally. A local server runs with your file system and your network, which is why it is priced without metering.

What you approve

5 tools, with what each one reads, writes and reaches shown before you agree — the same list on every route above. Read the tool surface.

Tool surface

What the model actually sees. Descriptions are diffed on every release — see version history.

Only tools that write or reach the network

Read from its source code. A server that builds its tool list when it starts can list more tools than this.

Ask Codex (OpenAI) to act as the planner: produce an implementation plan, review a design, or give a second opinion.

Parameter	Type	Required	Notes
request	string	yes	What to plan or review. Include goals, constraints and acceptance criteria.
cwd	string	yes	Absolute path of the project directory Codex should inspect.
context	string	no	Optional extra context: relevant diffs, error output, decisions already made.
session_id	string	no	Optional. Continue an earlier Codex thread instead of starting fresh.
effort	string	no	Optional reasoning effort. Default: the user's Codex config (high). Use medium/low for quick questions; xhigh/max for hard architecture problems.
model	string	no	Optional Codex model override. Default: the user's Codex config.
background		no	—

Ask Codex to generate (or edit) raster images with its image generation tool. Images are saved into output_dir as they are made, and their paths returned.

Parameter	Type	Required	Notes
prompt	string	yes	Description of the image(s): subject, style, size/aspect ratio, background (e.g. transparent), variations.
output_dir	string	yes	Absolute directory where the images should be saved (created if missing).
count	integer	no	Optional number of images wanted. Enables progress (k/count) and precise resumes.
filename	string	no	Optional base filename, e.g. "hero-banner.png".
reference_images	array	no	Optional absolute paths of images to attach as references or to edit.
session_id	string	no	Optional. Continue an earlier Codex thread (e.g. "make it darker").
effort	string	no	Optional. Default high; rarely needs changing.
model	string	no	Optional. Default gpt-6-luna; rarely needs changing.
background		no	—

Get the status or result of a Codex job (jobs persist across Claude sessions and restarts). With job_id: the result if finished, otherwise progress.

Parameter	Type	Required	Notes
job_id	string	no	Job id returned by codex_plan/codex_image.
wait_seconds	number	no	Optional. Seconds to wait for completion (0-…). Default 0.

Continue a job that was interrupted (reboot/crash), timed out or failed, from where it stopped: it continues the same Codex thread,

Parameter	Type	Required	Notes
job_id	string	yes	Job id to resume.

Stop a running Codex job.

Parameter	Type	Required	Notes
job_id	string	yes	Job id to cancel.

Free

$0 forever

Every tool, no call limit
No card, and mcprush charges nothing for it
Source published under MIT
Runs on your machine — nothing of it reaches our gateway
Nothing to cap, because mcprush bills nothing

Nothing here is billable: mcprush charges nothing to install Codex Bridge or to call it, at any volume.

ATrust score 93 / 100

Release 0.1.0 · scanned 28 Oct 2026

Two independent axes, because powerful and malicious are different questions. The grade is threat only. The capability level is blast radius, and it is never a penalty on the grade — it is priced as one subtract-only term in the score, where you can see it.

Threat score100

Client score93

Findings4 threat · 0 capability

Capabilityhigh

Coveragesource

Verificationrepo

Surface digestd59f11a4a70c9caff611ca1a0b0a8d0c582f4b67d08e8da93d0334dfd8b4d9cb

Score vectorthreat 100 → client 93

Term	Level	What it prices	Points
capability-exposure	high	capability blast radius (high) — client exposure if the model is manipulated	−6
verification-discount	repo	publisher verification (public source) — no provenance, but the source is public and inspectable	−1
coverage-honesty	source	inspection depth (source) — how much of the target the scan could see	−0

score = clamp(round(threatScore − E_cap − E_ver − E_cov), 0, 100). No model, no opinion, no network call at scoring time: the same methodology version against the same bytes always produces this number, and every line above is a field of the report rather than a summary of it.

What the scan could actually read

A grade is only as meaningful as its coverage, so the scanner publishes its own depth before it publishes its result.

Tool surfaceread

Implementation sourceread

Package metadataread

Live transportnot available

Tools were statically extracted from the published source (5 recovered), not enumerated from a running server. Tool-poisoning, Unicode-smuggling, capability and toxic-flow analysis ran on this inferred surface, but a mis-parsed registration could be missed or mis-attributed, so tool-derived findings are capped below “confirmed”. To grade the real runtime surface, scan the running server: --command "npx -y <package>".

Capability — what it could do if the model were manipulated

Tags derived from each tool’s schema and the implementation, not from what the tool calls itself. high is the level these add up to.

Tool	Capability tags	Why the tag was assigned
codex_plan	no tags
codex_image	no tags
codex_result	no tags
codex_resume	no tags
codex_cancel	no tags

Toxic-flow graph

The lethal trifecta, checked as a graph rather than as a checklist: untrusted input, a sensitive source and an external sink have to meet before there is a path worth worrying about.

The public result for this release does not print the flow graph, so there is nothing to show here. That is not the same as "no paths were found": what the scan did read is above, under coverage.

Supply chain and provenance

Integrityfirst-seen

Surface digestd59f11a4a70c9caf

Scanned2026-10-28

This is the first scan of this surface here, so there is nothing yet to compare it against.

Every result on this tab comes from one deterministic pass over the published package — offline, rule by rule, and auditable line by line above. Same methodology version, same bytes, same score.