mcp · United States

agentic-swmm-workflow

Agentic SWMM is an automated, auditable, and memory-informed framework for reproducible stormwater modelling, integrating QGIS and EPA SWMM through the aiswmm runtime, reusable Skills, and MCP interfaces, with QA verification, provenance tracking, calibration support, and Codex, Hermes, Claude code as well as OpenClaw compatibility

Visit source ↗GitHub repository ↗

Feature or claim this listing

Feature this listing — founding Featured slots are 50% off for 6 months

Represent agentic-swmm-workflow? Send a claim for review so we can discuss listing updates. Claims do not automatically grant ownership or verification.

Explore paid sponsorship →

What's in it

Data offered docs-verified: SWMM input files (model.inp) · SWMM report and binary outputs (.rpt, .out) · run manifests, command traces and QA summaries · parsed peak-flow metrics · rainfall-runoff figures · calibration summaries · fuzzy uncertainty summaries · audit records (experiment_provenance.json, comparison.json, experiment_note.md) · modelling memory summaries · SWMM networks from SWMMCanada (Canada) or SWMManywhere synthesis (elsewhere)
Coverage docs-verified: Canada: real storm networks for 35 cities, synthesis elsewhere in Canada; SWMManywhere-based synthesis outside Canada
Formats docs-verified: SWMM INP · SWMM RPT · SWMM OUT (binary) · JSON · GeoJSON · Parquet · PNG · Word (.docx) · Markdown (Obsidian-ready notes)
Access docs-verified: Free, open-source under MIT licence; local install via one-line installer, Docker image or pip (aiswmm); LLM provider API key needed for most routes, three of ten need none

Useful for

Use cases llm-inferred: Stormwater drainage model build from GIS inputs with QA and audit trail · Peak-flow and rainfall-runoff screening for drainage strategy options · Reproducible SWMM run handover with provenance records for client review
Audience llm-inferred: drainage engineers · hydrologists · stormwater modellers · GIS analysts

Limitations

Constraints llm-inferred: Pre-1.0 (v0.9.4), so interfaces may change · Real storm networks only for 35 Canadian cities via SWMMCanada; elsewhere SWMManywhere synthesis, which is not a calibrated or validated network · Validation rows are evidence-bounded: GIS preprocessing concept, prepared-input path and uncertainty smoke are not calibrated performance claims · Raw GIS path handles structured layers only, not arbitrary CAD/GIS recognition · ENTRY lists region as United States, but DOCS do not describe US or UK coverage, so UK suitability is unverified · Agent features need an LLM provider; one-line installers run a remote script · ENTRY auth says no credentials documented, but DOCS describe API key configuration · No MCP tool names or transport documented in DOCS; ENTRY tools list is empty · One evidence row (v0.7.1) states calibration is next-milestone scope

Licence: MIT
Authentication: No credentials documented in README
GitHub stars: 30
Last commit: 26/09/2026
Last successful page check: 14/10/2026
Page health: healthy
Package: Not recorded
Verification: Not independently verified

Provenance & evidence

Geospatial deep dive (Glama/Smithery/GitHub). Tools/install/auth auto-extracted from README.

Metadata observed 16/10/2026 · research sweep: geo-deep-dive (geo-deep-dive:github-search-geo)