con un clic
tscg-openclaw
TSCG Tool-Schema Compression for OpenClaw agents
Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.
Menú
TSCG Tool-Schema Compression for OpenClaw agents
Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.
Basado en la clasificación ocupacional SOC
| name | @tscg/openclaw |
| version | 1.4.1 |
| description | TSCG Tool-Schema Compression for OpenClaw agents |
| author | TSCG Team |
| license | MIT |
| tags | ["tscg","tool-compression","openclaw","optimization"] |
| minHostVersion | 2026.0.0 |
You are a TSCG-enhanced OpenClaw agent.
Your tool definitions are automatically compressed using TSCG (Tool-Schema Compression Grammar). This reduces token usage by 40-65% with <2pp accuracy impact.
The @tscg/openclaw plugin handles compression transparently via the beforeToolsList hook. If you notice unusual tool behavior, the original uncompressed definitions are always preserved as fallback.
Your model is matched to an optimal compression profile using a 4-tier resolution strategy:
| Model | Archetype |
|---|---|
| claude-opus | hungry |
| claude-sonnet | robust |
| claude-haiku | balanced |
| gpt-4 | sensitive |
| gpt-5 | robust |
| qwen3 / phi4 / llama3.1 / gemma3 / mistral | small-model |
| deepseek-v3 / deepseek-r1 | conservative |
For best results, run tscg-openclaw tune --model <your-model> to create a custom profile tailored to your specific model's compression tolerance.
Reference documents are available in the references/ directory:
self-tuning.md -- Guide to quick vs full benchmarks and cost optimizationbenchmarks.md -- 720-call benchmark results and operator impact datamodel-profiles.md -- Static profile map and size heuristic detailstroubleshooting.md -- Common issues and fixes