FeedbackJar
MCP Integration ✓ 2 minutes setup

FeedbackJar + OpenAI Codex

~/.codex/config.toml or .codex/config.toml

OpenAI Codex uses TOML config at ~/.codex/config.toml (or .codex/config.toml per project). Add FeedbackJar as an HTTP MCP server and Codex can triage your feedback backlog from the CLI or IDE.

Create API token → API tokens guide →

How to connect FeedbackJar to OpenAI Codex

1

Create an API token

bash 
# Workspace token: fj_live_...
# Export it so Codex can read it:
export FEEDBACKJAR_API_KEY="fj_live_..."

→ Codex does not inherit your full shell environment — use env_vars or bearer_token_env_var to forward secrets.

2

Option A — CLI

bash 
codex mcp add feedbackjar \
  --url https://api.feedbackjar.com/mcp \
  --bearer-token-env-var FEEDBACKJAR_API_KEY
3

Option B — config.toml

toml 
[mcp_servers.feedbackjar]
url = "https://api.feedbackjar.com/mcp"
bearer_token_env_var = "FEEDBACKJAR_API_KEY"
enabled = true

# Or with explicit headers:
# [mcp_servers.feedbackjar.http_headers]
# Authorization = "Bearer your_key_here"

→ Edit ~/.codex/config.toml, or create .codex/config.toml in a trusted project for project-scoped config.

4

Verify and use

bash 
codex mcp list

# Then in a Codex session:
# "What are the top open bugs in FeedbackJar?"

→ In the IDE extension, open MCP settings → Open config.toml from the gear menu.

What you get

  • ✓ TOML-based config (~/.codex/config.toml)
  • ✓ CLI: codex mcp add for quick setup
  • ✓ HTTP transport with bearer token auth
  • ✓ Project-scoped .codex/config.toml
  • ✓ Tool timeout and approval mode controls
  • ✓ codex mcp list to verify servers

Close the loop from OpenAI Codex

Users report bugs through your widget. Your agent reads the top votes via MCP, ships the fix, and every voter gets notified.

Start Free Trial 7-day free trial · No credit card required

Frequently Asked Questions

Looking for the full tool reference? Read the MCP Server docs →

Related

Widget integrations → All MCP agents → Compare alternatives →