Tessl Skills Evaluation Dataset for Agentic AI Tasks

A dataset accompanying the paper 'A Framework for Evaluating Agentic Skills at Scale'. It supports rigorous, reusable evaluation of agent skills derived from real-world challenges sourced from open-source GitHub repositories of trusted organizations. The dataset was created by author 'tesslio' and was last updated on June 5, -2026.

Use Cases

• Benchmarking agent performance on realistic, end-to-end challenges based on the described task structure.
• Developing reusable evaluation artifacts for specific agent skills based on the described knowledge artifacts.
• Comparing skill acquisition methods across different agent architectures based on the standardized task framework.

Strengths

• Tasks are derived from real-world skills sourced from trusted, well-known organizations' GitHub repositories.
• Designed for rigorous, reusable evaluation as part of a published academic framework.
• Last update recorded as 2026-06-05, suggesting recent maintenance.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and license information are unknown, which may limit suitability assessment.

Provenance

Source

Open-source GitHub repositories from trusted, well-known organizations.

Collection Method

Tasks derived from real-world skills or groups of related skills.

Time Range

null

Freshness

Last updated 2026-06-05 14:34:04; freshness should be verified.

Geography

null