Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@genpark-agent-synthetic-trajectory-evaluator-skill Score my agent's trajectory against the gold benchmark with precision, recall, and edit distance."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
genpark-agent-synthetic-trajectory-evaluator-skill delivers zero-dependency, low-latency web automation, DOM semantic pruning, and execution trajectory evaluation primitives engineered strictly using Python 3.9+ standard library.
Zero External Dependencies: Operates exclusively via pure Python (html.parser, urllib.parse, re, math, json). Zero pip install overhead, zero headless browser crashes.
Enterprise Web Agent Invariants: Implements formal token-pruning algorithms, form auto-mapping, anti-crawler trap normalization, Markdown-to-JSON type inference, and trajectory Levenshtein distance evaluation.
Native Anthropic MCP Protocol: Compliant with standard JSON-RPC 2.0 stdio MCP specifications for Claude Desktop, Cursor, and Windsurf.
Related MCP server: genpark-web-form-input-schema-auto-mapper-skill
flowchart TD
RawWeb["Raw Web Page / DOM Ingress"] --> TrapFilter["URL Canonicalization & Anti-Crawler Trap Guard"]
TrapFilter --> DOMPruner["HTML DOM Semantic Tree Pruner
(80%+ Token Reduction, Strips Scripts/Styles/SVG)"]
DOMPruner --> FormMapper["Web Form Input Schema Auto-Mapper
(Attribute & Heuristic Profile Field Binding)"]
DOMPruner --> TableParser["Markdown & HTML Table to JSON Transformer
(Type-Inferred Structured Record Generation)"]
FormMapper --> AgentExecution["Autonomous Agent Browser Interaction"]
TableParser --> AgentExecution
AgentExecution --> TrajectoryEval["Synthetic Trajectory Evaluator
(Action Precision, Recall & Levenshtein Edit Distance)"]
TrajectoryEval --> VerifiedTaskDone["Verified Benchmark Task Completion"]from client import AgentSyntheticTrajectoryEvaluator
# Initialize engine
engine = AgentSyntheticTrajectoryEvaluator()
# Execute self-testing benchmark suite
result = engine.run_benchmark_trajectory_evaluator()
print("Execution Result:", result)Add to your claude_desktop_config.json or cursor.json:
{
"mcpServers": {
"genpark-agent-synthetic-trajectory-evaluator-skill": {
"command": "python",
"args": ["-u", "/path/to/genpark-agent-synthetic-trajectory-evaluator-skill/mcp_server.py"]
}
}
}This skill contains pre-configured smithery.yaml and pyproject.toml manifests. Install directly via pip:
pip install git+https://github.com/alphaparkinc/genpark-agent-synthetic-trajectory-evaluator-skill.gitThis server cannot be deployed
Deterministic web intake and data utilities for autonomous agents.
Turn any webpage into a structured action manifest — clickable, fillable, submittable elements.
Clean Markdown and AI-readability scoring for any URL. Built for AI agents.
Headless browser primitives for AI agents when sites need real JS rendering.
