A tool for running a HEC-RAS computer model without its normal graphical interface and checking the results. HEC-RAS is software used to model water flow and flooding in rivers and channels.

Use it to copy and run a HEC-RAS project and plan, accept the software's terms prompt when needed, enforce a timeout, and decide whether the computed results are usable.

It provides a repeatable run and quality check, including warnings, errors, water-surface results, volume differences, output files, and runtime details.

ras-runner — Skill for Claude Code and Codex by Zhonghao1995

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent

npx skills add Zhonghao1995/Agentic-HEC-RAS --skill ras-runner

Clone the repo

git clone --depth 1 https://github.com/Zhonghao1995/Agentic-HEC-RAS

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/zhonghao1995/agentic-hec-ras/ras-runner.svg)](https://agentmods.dev/skills/zhonghao1995/agentic-hec-ras/ras-runner)

Your own site

<a href="https://agentmods.dev/skills/zhonghao1995/agentic-hec-ras/ras-runner"><img src="https://agentmods.dev/badge/skills/zhonghao1995/agentic-hec-ras/ras-runner.svg" alt="Measured on agentmods" height="20"></a>

Per session 128 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 955 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00128	$0.00955
Opus 5	$0.00064	$0.00477
Sonnet 5	$0.00026	$0.00191
Haiku 4.5	$0.00013	$0.00096

Measured 6d ago against content hash 1ef3cbb3c343, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

ras-runner scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/ras-runner/SKILL.md · 40 lines

How it starts

The opening of the file, as written. The whole thing — 40 lines — stays where its author put it; the contents beside it link to each section on GitHub.

ras-runner: headless run + QA gate

Tool

ras_run(project, plan, run_dir?, num_cores?, overwrite?, clear_geompre?, timeout_s?, accept_tcu?, ras_exe?)

runs Ras.exe -c <project.prj> <project.p##> (geometry preprocessing + unsteady/steady compute + post-processing), blocking
run_dir given -> the project is copied there first and run inside it (safe); omitted -> runs in place (only on a copy!)
a watchdog thread handles HEC-RAS's Terms and Conditions for Use (TCU) form (accept_tcu, default true: pages through it and presses "I agree" / OK; false: terminates the run and reports failed) and kills the run after timeout_s (default 14400 s); what it did is listed in watchdog_events
returns status, usable, reasons, metrics{max_wsel_error, volume_error_pct, n_error_lines, n_warning_lines}, hdf (the .p##.hdf), runtime (unsteady speed hr/hr, complete process hr), elapsed_s, log_file, log_tail, watchdog_events, ras_exe / ras_version

Procedure

Copy (ras_project_copy) and edit first (ras-model).
ras_run(project=<copy>, plan="03", num_cores=4).
Gate: usable = status in {completed, completed_with_warnings} AND hdf non-null.
- completed_with_warnings: usable, but quote reasons (max WSEL error > 0.05, |volume error| > 1 %, WARNING lines).
- unstable / failed / no_results / unknown: NOT usable — report reasons + log_tail, do not read results, stop the branch.
Hand hdf to ras-results; keep the whole ras_run return for ras_manifest_write(run=...).

Rules

HEC-RAS is auto-detected (HECRAS_EXE env var, <repo>/.local/HEC-RAS/<ver>/Ras.exe, C:/Program Files (x86)/HEC/HEC-RAS/<ver>/Ras.exe, newest first). If ras_run returns ok:false mentioning HECRAS_EXE, HEC-RAS is not installed: stop and say so (results of earlier runs can still be read).
Runs block until HEC-RAS exits; a 2D run can take minutes to hours. Do not run the same copy twice concurrently; run baseline and variant sequentially or in separate copies.
Runtime scales with cores only up to ~4-8; num_cores overrides the plan's setting for this run.
After geometry edits (not supported by these tools) pass clear_geompre=true; otherwise leave it.
HEC-RAS 7.0.1 can show its "Terms and Conditions for Use" form even on headless launches (seen on some, not all, runs). The built-in watchdog accepts it (~15 s overhead) unless accept_tcu=false; if a run still returns failed with a TCU reason, ask the user to open HEC-RAS once and accept.
Set timeout_s to something realistic for big 2D models; a timed-out run is killed and returns status=failed with the reason.
Linux compute engines exist (RasUnsteady, .local/HEC-RAS/7.0.1/Linux) but need Windows-side preprocessing first; not wired into ras_run in v0.1.

Read the full file on GitHub · 40 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

6d ago First seen · 40 lines · 128 tokens per session scan A 1ef3cbb3c343

Subscribe to this mod's changes

ras-runner is a skill published in the GitHub repository Zhonghao1995/Agentic-HEC-RAS (1 stars, last pushed 21d ago), licensed MIT. It adds 128 tokens to every session and 955 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Claude Code Scientific agentic-ai flood-modelling hec-ras hydraulic-modeling mcp

mike-end-to-end

Standard operating procedure for an auditable, headless MIKE+ workflow via the mike-plus MCP server — how to plan, copy, run, QA-gate, compare, plot, audit, and safely stop. Use FIRST whenever an agent is handed a MIKE+ model (.sqlite) or result (.res1d) plus a plain-language goal; it picks the operating mode and…

Zhonghao1995/Agentic-MIKE-Plus · 98 tokens

mike-compare

Compare two MIKE+ runs via the mike-plus MCP server — baseline vs scenario, before vs after a parameter edit — to say WHERE a change mattered and BY HOW MUCH (delta peak, delta volume, timing shift, RMSE/NSE) and to draw the overlay hydrograph. Use whenever two .res1d files of the same model exist or a re-run just…

Zhonghao1995/Agentic-MIKE-Plus · 93 tokens

mike-plot

Produce publication-style figures from MIKE+ .res1d results via the mike-plus MCP server — a stacked rainfall-runoff hydrograph, a single time series, a network map, a baseline-vs-scenario overlay, or a longitudinal profile (bed / crown / ground / max water level along a pipe string), in the house style (Arial 12…

Zhonghao1995/Agentic-MIKE-Plus · 109 tokens

mike-results

Read MIKE+ .res1d results via the mike-plus MCP server — list contents, summarize peaks, extract a single time series, or check which nodes flood (peak water level vs ground level). Use when an agent needs to inspect or pull numbers from MIKE+ network results (node water levels, link/reach discharge, velocities…

Zhonghao1995/Agentic-MIKE-Plus · 84 tokens

mike-audit

Record provenance for an Agentic MIKE+ step via the mike-plus MCP server — a manifest.json tying the model copy, inputs, edits, run QA status, result files and figures together, each file with a sha256. Use at the end of every run/compare/plot task (successful or failed) so the work can be audited, re-compared or…

Zhonghao1995/Agentic-MIKE-Plus · 86 tokens

mike-import

Bring an EPA SWMM (.inp) or EPANET (.inp) model into MIKE+ via the mike-plus MCP server — the bridge from an INP builder such as SWMMCanada or agentic-swmm into a MIKE+ database that mikemodelinfo / mikerun can use. Use when the user has an .inp and wants it in MIKE+. Requires MIKE+ installed + a checked-out license…

Zhonghao1995/Agentic-MIKE-Plus · 97 tokens