NEPA MCP v0.1.2

NEPA MCP v0.1.2 updates PAD-US protected-area screening and improves installation guidance.

Highlights

• Queries the USGS PAD-US 4.1 Combined layer so designation, easement, marine, proclamation, ownership, and management records can surface in the general protected-area tool.
• Expands PAD-US owner codes into readable labels and reports record counts by owner type and category.
• Identifies GIS_Acres as the full mapped acreage of each intersecting source feature. These overlapping values are not additive and do not represent total land area within the ROI.
• Improves pipx recovery guidance and PyPI visibility in the README.

Thanks to @adlerarcher for initiating the PAD-US Combined-layer correction in #17.

Validation

• 1,129 tests passed (1 skipped)
• Ruff formatting and lint checks passed
• Python 3.12 and Python 3.14 CI passed
• Package build, metadata checks, clean-wheel installation, and all 19-server/46-tool stdio smoke tests passed
• PAD-US live validation passed at 13 locations
• Published through PyPI Trusted Publishing with attestations

Install or upgrade from PyPI.

NEPA MCP v0.1.1

NEPA MCP v0.1.1 improves installation and MCP client configuration.

Highlights

• Resolves the installed nepa-mcp launcher when generating Claude Code, VS Code, and Codex configuration, avoiding GUI PATH issues and stale global launchers.
• Simplifies PyPI installation, upgrades, and MCP client setup guidance.
• Clarifies the Codex CLI and Codex Desktop plugin paths.
• Clarifies Census and EPA AQS credential requirements.

Validation

• 1,127 tests passed (1 skipped)
• Clean installs from PyPI on Python 3.12 and 3.14
• All 19 MCP servers and 46 tools verified over stdio on both Python versions
• Published through PyPI Trusted Publishing with attestations

Install or upgrade from PyPI.

NEPA MCP v0.1.0

NEPA MCP v0.1.0

First stable public release of the PermitAI NEPA MCP Toolkit.

Included

• 19 MCP servers and 46 tools for federal environmental data, regulatory research, and geospatial screening.
• Interactive map composition across 32 selectable environmental layers.
• Production installation from PyPI with pipx.
• Client configuration generation for Codex, Claude Code, and VS Code.
• A Codex plugin and nepa-screening skill installable from this repository marketplace.

Install

pipx install nepa-mcp
nepa-mcp doctor
nepa-mcp list-servers

Python 3.12 or newer is required.

For Codex Desktop, add pnnl/nepa-mcp as a plugin marketplace with Git ref v0.1.0, leave Sparse paths blank, then install NEPA-MCP. The Python runtime must be installed first using the command above.

Validation

• 1,124 tests passed and 1 skipped.
• Ruff, packaging, and dependency-audit checks passed.
• Clean production PyPI installs passed on Python 3.12 and 3.14.
• All 19 MCP servers started over stdio and listed 46 tools.
• Representative live GIS and eCFR calls passed.
• A fresh Codex marketplace and plugin install from Git ref v0.1.0 passed.

Artifact SHA-256:

• Wheel: 21aa2955f75572024c768c2bd209850f647e9a96ea49e5876dd8d993c0b534ce
• Source distribution: 6927bb9d1dd00a3eb4a9ccd168eab131a59552b3630c32b09319725024de21c4

NEPA MCP is a screening and research aid. It does not make legal or agency determinations, replace consultation with agencies or Tribes, or guarantee that an upstream dataset is complete or current.

• Documentation: https://pnnl.github.io/nepa-mcp/
• PyPI: https://pypi.org/project/nepa-mcp/0.1.0/

NEPA MCP v0.1.0rc2

NEPA MCP v0.1.0rc2

Second public release candidate of the PermitAI NEPA MCP Toolkit.

Changes since rc1

• Records the package source validated through TestPyPI.
• Makes README images and notices render correctly on package indexes.
• Adds direct Documentation and Releases project links.
• Keeps repository-based installation as the public installation route until the production PyPI release.

Validation

• 1,124 tests passed and 1 skipped.
• Ruff and distribution checks passed.
• No known vulnerabilities were found in resolved dependencies.
• Clean package installs passed on Python 3.12 and 3.14.
• All 19 MCP servers started over stdio and listed 46 tools.
• Representative GIS and CFR calls passed.

TestPyPI candidate: https://test.pypi.org/project/nepa-mcp/0.1.0rc2/

Artifact SHA-256:

• Wheel: f27cadbeabe1159c6a894ad336442abf73eb789299ab2ef844bfc25ff9c9ee1e
• Source distribution: 50458e4dd47edb25bb3e486a41412dff63bc4f83af38072e03004d0501840e10

Install from the repository

git clone https://github.com/pnnl/nepa-mcp.git
cd nepa-mcp
pipx install .
nepa-mcp doctor
nepa-mcp list-servers

Release status

This is a pre-release. Production PyPI publication is not part of this release.

NEPA MCP is a screening and research aid. It does not make legal or agency determinations, replace consultation with agencies or Tribes, or guarantee that an upstream dataset is complete or current.

Documentation: https://pnnl.github.io/nepa-mcp/

NEPA MCP v0.1.0rc1

NEPA MCP v0.1.0rc1

First public release candidate of the PermitAI NEPA MCP Toolkit.

Included

• 19 MCP servers and 46 tools for federal environmental data, regulatory research, and geospatial screening.
• Repository-based installation with pipx.
• Client configuration generation for Codex, Claude Code, and VS Code.
• Optional Codex plugin installation from the repository marketplace.
• Public documentation through GitHub Pages.

Install

git clone https://github.com/pnnl/nepa-mcp.git
cd nepa-mcp
pipx install .
nepa-mcp doctor
nepa-mcp list-servers

Release status

This is a pre-release. Installation is currently from the GitHub repository;
PyPI publication is not part of this release. Census and EPA AQS integrations
accept optional credentials, while the other 17 servers are credential-free.

NEPA MCP is a screening and research aid. It does not make legal or agency
determinations, replace consultation with agencies or Tribes, or guarantee
that an upstream dataset is complete or current.

Documentation: https://pnnl.github.io/nepa-mcp/