Skip to main content

JD Codec — privacy-first local MCP proxy for browser-agent tasks. Compresses page snapshots ~80%+ via a cloud codec; PII never leaves your machine.

Reason this release was yanked:

bug fix

Project description

jdcodec — Python Distribution

pip install jdcodec

A privacy-first local proxy for browser-agent task pipelines. The Python distribution is a thin wrapper that delegates to the JD Codec connector — it gives you pip install jdcodec ergonomics with the connector running under the hood via Node.

What it does

  • Compresses browser-agent page snapshots so your agent's context window goes much further on long tasks.
  • Scans every snapshot on your machine for PII (emails, phone numbers, credit cards, API keys, addresses, and more) and replaces them with category tokens before any bytes leave your network.
  • Falls back to a local-only mode if the cloud service is unreachable — your agent never blocks on us.

Requirements

  • Python 3.8 or newer
  • Node.js 22 or newer on your PATH (install from https://nodejs.org/ or brew install node / apt install nodejs npm)
  • An MCP-capable agent client (Claude Code, Cursor, or any client that speaks MCP stdio)
  • A JD Codec API key — email hello@jdcodec.com

Usage

The jdcodec command exposes three surfaces:

pip install jdcodec
jdcodec                 # launches the local MCP proxy — point your
                        #   MCP-capable agent client at this command
jdcodec audit           # one-shot environment audit (detects langchain /
                        #   playwright / browser-use, posts a telemetry
                        #   ping). `jdcodec-audit` is an alias.
jdcodec start           # registers your machine for the alpha waitlist —
                        #   opens browser-based OAuth, captures email,
                        #   attaches it to your audit. `jdcodec login`
                        #   is an alias.

The first MCP-proxy invocation fetches and caches the connector via npx (a few seconds; instant on subsequent runs).

Pinning a specific connector version

By default the wrapper resolves to the latest published connector. To pin:

JDC_CONNECTOR_VERSION=0.3.1 jdcodec

Version pinning only applies to the MCP-proxy path (which shells through to npm). The audit / start / login subcommands are local Python and don't read this variable.

Wiring it into your agent

The Python wrapper exposes the same jdcodec command as the npm distribution. Point your MCP client config at jdcodec (not npx @playwright/mcp) and the connector takes over. Specific guides for Claude Code, Cursor, and generic MCP clients are available — email hello@jdcodec.com for access.

Configuration

Set your API key either as an environment variable or in a config file:

export JDC_API_KEY='jdck_yourid.yoursecret'

Or:

mkdir -p ~/.jdcodec
cat > ~/.jdcodec/config.json <<EOF
{ "api_key": "jdck_yourid.yoursecret" }
EOF
chmod 600 ~/.jdcodec/config.json

The full env-var reference (JDC_BYPASS, JDC_REGION, JDC_TRACE, etc.) is documented in the npm package's README — npm view jdcodec readme or visit https://www.npmjs.com/package/jdcodec.

Support

Email hello@jdcodec.com for alpha access, API keys, bug reports, or feature requests.

Project details


Release history Release notifications | RSS feed

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

jdcodec-0.4.0.tar.gz (14.9 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

jdcodec-0.4.0-py3-none-any.whl (11.6 kB view details)

Uploaded Python 3

File details

Details for the file jdcodec-0.4.0.tar.gz.

File metadata

  • Download URL: jdcodec-0.4.0.tar.gz
  • Upload date:
  • Size: 14.9 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.13

File hashes

Hashes for jdcodec-0.4.0.tar.gz
Algorithm Hash digest
SHA256 51defc9af215783e64ef12bffc72933177198f93eb40c7dac993213c104fcffc
MD5 876ba2f85f2f3ecfd9c01dfef0021ffc
BLAKE2b-256 1005376e6c448585433bf455ece9a3cdcf445527c37dad13b31e7569de8b98b7

See more details on using hashes here.

File details

Details for the file jdcodec-0.4.0-py3-none-any.whl.

File metadata

  • Download URL: jdcodec-0.4.0-py3-none-any.whl
  • Upload date:
  • Size: 11.6 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.13

File hashes

Hashes for jdcodec-0.4.0-py3-none-any.whl
Algorithm Hash digest
SHA256 a27a3b54baba3dba1d8e42699343358aeeff08b6dc9f665fceb9d667d2e28c3a
MD5 0e4c70d86d3477cfbfb1a8d88e94f563
BLAKE2b-256 4f16677e84975e975e107d988a5390a0df25ef374a584c16c962787697957169

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page