Install
$ agentstack add mcp-17557295964-ops-project-brain-bridge ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Project Brain Bridge / 项目大脑桥接
Project Brain Bridge is a reusable Codex skill for connecting a local project to ChatGPT as a review and planning brain.
它解决的问题很明确:让 ChatGPT 读取项目状态、提出阶段计划,让 Codex 在本地执行、回写证据,再让 ChatGPT 复审。它不是某个具体项目的脚本,也不绑定任何私人路径。
What It Provides
- A reusable
.ai-bridgeproject state structure. - A read-only MCP bridge for project status review.
- A narrow write-stage MCP bridge for writing
current-plan.md. - Prompt templates for read-only diagnosis, stage planning, and review.
- Local setup and doctor scripts.
- Security boundaries that keep secrets, tokens, raw private data, and unrelated projects out of scope.
When To Use It
Use this skill when you want:
- ChatGPT to act as a skeptical project brain.
- Codex to act as the local executor.
- A repeatable loop:
state -> plan -> execution -> status/log/diff -> review. - A bridge pattern that can be reused across different local projects.
Do not use it to give ChatGPT unrestricted filesystem write access. The write bridge is intentionally narrow.
Install As A Codex Skill
Clone this repository directly into your Codex skills folder:
mkdir -p ~/.codex/skills
git clone https://github.com/YOUR_ACCOUNT/project-brain-bridge.git ~/.codex/skills/project-brain-bridge
Then invoke it in Codex:
$project-brain-bridge
If you are using the repository locally before publishing:
cp -R /path/to/project-brain-bridge ~/.codex/skills/project-brain-bridge
Quick Start
Initialize bridge files inside a target project:
python3 ~/.codex/skills/project-brain-bridge/scripts/init_ai_bridge.py /path/to/your/project
Start or adapt the read-only MCP server from:
assets/server-template/readonly_bridge_server.js
Start or adapt the write-stage MCP server from:
assets/server-template/write_stage_bridge_server.js
Run the doctor:
python3 ~/.codex/skills/project-brain-bridge/scripts/doctor.py \
--project-root /path/to/your/project \
--readonly-url http://127.0.0.1:18891/mcp \
--write-url http://127.0.0.1:18893/handoff-mcp-v2 \
--write-test
Recommended Workflow
- Codex initializes
.ai-bridge. - ChatGPT uses the read-only connector to inspect state and evidence.
- ChatGPT writes a bounded stage plan through the write connector.
- Codex reads
current-plan.mdand executes locally. - Codex writes status, log, and diff evidence.
- ChatGPT reviews the result before the next stage.
Security Model
The default design separates reading and writing:
- Read-only connector: project state inspection only.
- Write-stage connector: writes only
.ai-bridge/current-plan.md. - No secret reading by default.
- No direct source-code modification by ChatGPT.
- No broad filesystem write tool.
Before using this with a real project, review:
references/architecture.mdreferences/deployment.mdreferences/mode-workflow.mdreferences/prompts.md
Repository Structure
SKILL.md
agents/openai.yaml
assets/
launchagent-template/
server-template/
starter-ai-bridge/
references/
scripts/
License
MIT. See LICENSE.
Source & license
This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: 17557295964-ops
- Source: 17557295964-ops/project-brain-bridge
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.