ChatGPT to Local BridgeBLOCK
Codex/ChatGPT plugin app and OAuth MCP connector for approved local workspaces.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
ChatGPT2LocalBridge Codex / ChatGPT Plugin App for approved local workspaces.
Pages source · Showcase · Human tutorial · Agent tutorial · Auth modes · Linux deploy · Alternatives · Sync flows · Skill runtime · Handoff spec · Codex analytics · Evidence · Promo copy · Mobile promo · Roadmap
Linux One-Click Prompt
Repository: https://github.com/Harzva/chatgpt2localbridge
Copy t
9a9fd063221eOBSERVED · 2026-10-07Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add chatgpt2localbridge --env CODEX_ANALYTICS_API_KEY_ENV=${CODEX_ANALYTICS_API_KEY_ENV} -- npx -y [email protected]{
"mcpServers": {
"chatgpt2localbridge": {
"command": "npx",
"args": [
"-y",
"[email protected]"
],
"env": {
"CODEX_ANALYTICS_API_KEY_ENV": "${CODEX_ANALYTICS_API_KEY_ENV}"
}
}
}
}Exposed tools (71)
52 read · 18 write · 1 destructive. Blast radius: 1 tool can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.
| Tool | Risk | Description |
|---|---|---|
batch_read | read | |
bridge.activity | read | |
bridge.health | read | |
bridge.logs | read | |
bridge.status | read | |
bridge_health | read | |
cloud.download | read | |
code.read | read | |
code.read_range | read | |
code.search | read | |
codex.cancel | read | |
codex.result | read | |
codex.status | read | |
codex.task_start | write | |
codex_result | read | |
codex_status | read | |
codex_task_start | write | |
file.copy | read | |
file.delete | destructive | |
file.list | read | |
file.mkdir | read | |
file.move | write | |
file.patch | write | |
file.read_path | read | |
file.stat | read | |
file.write | write | |
file_list | read | |
file_read_path | read | |
file_write | write | |
git.checkpoint | read | |
git.diff | read | |
git.revert | read | |
git.status | read | |
handoff.create | write | |
handoff_create | write | |
local_bundle_dir | read | |
local_list_dir | read | |
local_read_file | read | |
local_workspace_action | read | |
local_write_file | write | |
policy.read | read | |
policy.validate | read | |
policy_read | read | |
port.check | read | |
process.list | read | |
process.start | write | |
process.stop | write | |
project.bundle | read | |
project.index | read | |
project.scripts | read | |
project.snapshot | read | |
service.restart | write | |
shell.exec | write | |
shell_exec | write | |
skill.bundle | read | |
skill.list | read | |
skill.read | read | |
skill.route | read | |
skill.search | read | |
task.finish | read | |
task.note | read | |
task.start | write | |
task.status | read | |
test.detect | read | |
test.run | write | |
trace.session_current | read | |
trace.session_end | read | |
trace.session_start | write | |
workspace.add | write | |
workspace.list | read | |
workspace.resolve | read |
Trust audit
BLOCKgrade D · trust 69/100 Do not install this without reading the findings. The audit found something that could harm you or your machine.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | FAIL |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (9 observation(s))
- Network
- declared (10 observation(s))
- Shell
- declared (5 observation(s))
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (19)
" \"denyGlobs\": [\"**/.env\", \"**/.env.*\", \"**/*.pem\", \"**/*.key\", \"**/.ssh/**\"],\n",
"**/.ssh/**".to_string(),
"**/.netrc" => file_name == ".netrc",
"**/.ssh/**" => normalized.contains("/.ssh/") || normalized.ends_with("/.ssh"),"**/id_rsa" => file_name == "id_rsa",
file.delete
http://127.0.0.1:3842/app
-> http://127.0.0.1:3838/mcp
curl -sS http://127.0.0.1:3838/health
http://127.0.0.1:3838/app
LOCALBRIDGE_CODEX_BASE_URL=http://127.0.0.1:4999/v1
@modelcontextprotocol/sdk, zod, @types/node
assets/cover.png
docs/assets/cover.png
curl -fsSL https://raw.githubusercontent.com/Harzva/chatgpt2localbridge/main/scripts/linux-one-click-install.sh | bash
curl -fsSL https://raw.githubusercontent.com/Harzva/chatgpt2localbridge/main/scripts/linux-one-click-install.sh | bash
curl -fsSL https://raw.githubusercontent.com/Harzva/chatgpt2localbridge/main/scripts/linux-one-click-install.sh | bash
curl -fsSL https://raw.githubusercontent.com/Harzva/chatgpt2localbridge/main/scripts/linux-one-click-install.sh | bash
curl -fsSL https://raw.githubusercontent.com/Harzva/chatgpt2localbridge/main/scripts/linux-one-click-install.sh | bash
Gates applied: no_behavioural_pass.
9a9fd063221efull audit observations/trust-audit/mcp-server/harzva__chatgpt-to-local-bridge.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-07 | 9a9fd063221e | BLOCK | D | 69 | first audit |
Questions
What is the ChatGPT to Local Bridge MCP server?
Codex/ChatGPT plugin app and OAuth MCP connector for approved local workspaces.
What tools does ChatGPT to Local Bridge expose?
71 in total: 52 read-only, 18 that write, and 1 that can delete or overwrite (file.delete). Every one is listed on this page with its risk.
Is ChatGPT to Local Bridge safe to connect to an agent?
No — not without reading the findings first. The audit graded it D (69/100) and found 5 critical or high issues in the source. Each one is listed on this page with the file and line it is on. Separately from the audit: 1 of its tools can destroy data, so scope the token you give it to what you actually need.
What credentials does ChatGPT to Local Bridge need?
It reads CODEX_ANALYTICS_API_KEY_ENV from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does ChatGPT to Local Bridge run?
It speaks stdio and streamable-http, so it runs as a local process your client starts. It is published on npm as chatgpt2localbridge at 0.1.1.
How current is this page?
The grade is for one exact copy of the source (9a9fd063221e), read on 2026-10-07. The repository is watched and re-audited when it changes.