← 返回 Skills 市场

Fluid Network Agent

Name: Fluid Network Agent
Author: zebra6-web

作者 Zebra6-web · GitHub ↗ · v1.0.0 · MIT-0

cross-platform ✓ 安全检测通过

149

总下载

当前安装

版本数

在 OpenClaw 中安装

/install fluid-network-agent-lenovo

功能描述

从 TOML 读取稳态不可压缩流体网络，计算各节点压力与各管路流量/压差（可选速度），并按工况输出连通性与负载可靠性（PASS/FAIL）。当用户给出网络 TOML 或询问某工况下系统是否能正常工作时使用。

使用说明 (SKILL.md)

流体网络求解与可靠性报告

该 Skill 会从 TOML 里读取：

节点：junction（未知压力）和 reservoir（固定压力）
管路：阻力模型（linear 或 quadratic），可选的 gating 用于表示阀门/故障开关
工况：有限变量赋值（例如 valve_v1 = "open"/"closed"）
功能判据：由一个或多个负载节点组成的 PASS/FAIL 规则

然后对每个工况计算：

节点压力与各管路流量/压差（需要几何信息时可给出速度）
仅基于启用边的连通性，并按阈值判断负载/功能是否可靠（PASS/FAIL）

用户要做下面任意一件事时用这个 Skill：

计算“各节点压力 / 各管路流量/流速 / 系统流路”
提供流体网络 TOML，或要求把需求整理成 TOML
需要对多个阀门/故障工况做对比，并判断“哪个工况能让哪个负载/功能工作”
TOML 文件路径；至少包含 nodes、edges，scenarios / functions / analysis 可选但推荐。

如果用户没有提供 TOML，让 Agent 按 reference.md 的格式补全成可运行版本（参考 examples/example_network.toml）。

执行顺序

先确认用户想看什么：要不要按“工况”对比、以及每个负载/功能的阈值条件。
检查 TOML 字段是否齐全（缺什么就让用户补齐，或让 Agent 补齐最小可运行字段）。
跑计算：python scripts/run_fluid_network.py --input \x3Ctoml_path> --format markdown（需要指定工况就加 --scenario \x3Cscenario_id>）。
失败就按报错定位字段/引用问题，并最多重试两轮。
输出 Markdown 报告，按工况给出连通性、各边结果和功能 PASS/FAIL。

输出格式

# Fluid Network Scenario Report

## Scenario: \x3Cid>
### Connectivity
- Reachable nodes from sources: ...
### Edge results
- \x3Cedge_id>: enabled=..., Q=... m3/s, v=... m/s|n/a, dP=... Pa, R_eff=...
### Function reliability
- \x3Cfunction_id>: PASS|FAIL
  - Load \x3Cload_node_id>: PASS|FAIL, connected=..., p=... Pa, flow=... m3/s

约定与假设（输出时保持一致）

Steady-state, incompressible, network-level mass balance at junction nodes.
Flow direction uses edge orientation: positive Q is from -> to.
Linear model: dP = R * Q. Quadratic model: dP = R * Q^2.
If an edge is scenario-disabled by gating, the solver multiplies its resistance by a large factor to approximate a blocked pipe (and connectivity uses enabled=boolean).

安全使用建议

This skill appears internally consistent and implements the described solver/reporting flow. Before installing or running it: 1) Run it in an isolated environment (virtualenv/container). 2) Ensure Python 3.11 or the tomli package is available (the script tries tomllib then tomli). 3) Inspect and, if needed, test with the provided examples/example_network.toml so the agent cannot fabricate parameters silently — SKILL.md allows the agent to auto-complete missing TOML fields, so verify any generated values. 4) Note there are a few code-quality issues to check (the distributed source appears truncated in places and there is an obvious bug near the end of solver.py where a variable name looks incorrect — e.g., use of an undefined variable 'a' for area/velocity calculation); you may want the author to fix/release a corrected version before using on important data. 5) The skill does not access secrets or network resources; nevertheless, treat it as untrusted third-party code: inspect, run tests, and avoid providing any sensitive credentials in TOML or other inputs.

功能分析

Type: OpenClaw Skill Name: fluid-network-agent-lenovo Version: 1.0.0 The skill bundle provides a legitimate numerical solver for steady-state incompressible fluid networks, including linear and non-linear (Newton-Raphson) solving capabilities. The code in 'scripts/run_fluid_network.py' and the 'src/fluid_network/' directory is well-structured, implements standard engineering algorithms, and lacks any indicators of data exfiltration, malicious execution, or prompt injection. All file operations are restricted to reading user-provided TOML files and outputting reports as requested.

能力评估

✓ Purpose & Capability

Name/description match the included code and SKILL.md: the package parses TOML network descriptions, runs numeric solves, and emits connectivity / PASS‑FAIL reports. No unrelated binaries, env vars, or credentials are requested.

ℹ Instruction Scope

SKILL.md instructs the agent to validate/complete a TOML and run `python scripts/run_fluid_network.py --input <toml>` to produce a report. Allowing the Agent to '补齐最小可运行字段' (auto-complete missing TOML fields) is functionally reasonable for convenience, but it gives the agent discretion to fabricate parameter values if the user doesn't review/confirm them. Otherwise instructions limit actions to loading the TOML, running the solver, and returning results (no file reads beyond the TOML/output file).

✓ Install Mechanism

No install spec; code is provided in the skill bundle and execution relies on a local Python runtime. No external downloads or non-standard install paths are invoked. The code uses tomllib (py3.11+) or tomli fallback—this is a normal dependency but not declared as an install step.

✓ Credentials

The skill requires no environment variables, credentials, or config paths. All required inputs are user-provided TOML files and optional CLI flags, which are proportionate to the stated purpose.

✓ Persistence & Privilege

The skill does not request 'always: true' or any elevated/always-on presence. It does not attempt to modify other skills or system-wide agent settings in the provided files or instructions.

如何使用

确保已安装 OpenClaw（本地或 Docker 部署）
在对话框中输入安装命令：/install fluid-network-agent-lenovo
安装完成后，直接呼叫该 Skill 的名称或使用 /fluid-network-agent-lenovo 触发
根据 Skill 的参数说明提供必要输入，即可获得结构化输出

版本历史

v1.0.0

initial release

元数据

Slug fluid-network-agent-lenovo

版本 1.0.0

许可证 MIT-0

累计安装 0

当前安装数 0

历史版本数 1

常见问题