Abstract
A technique is proposed herein for converting artificial intelligence (AI)-generated network testcase intents and parameter combinations into detailed procedure structures through two independent reuse decisions. Technical-artifact applicability controls whether evidence may be shared, while execution compatibility controls whether an ordered procedure can be shared. A machine-readable row contract preserves testcase constraints for both decisions. One template is generated per compatible procedure family and is bound deterministically to row-specific values and a selected combination, retaining evidence authority, unresolved exact artifacts, and validation lineage. In a matched-input comparison of complete workflows, a working implementation reduced completed AI-agent turns, total tokens and input tokens not served from cache, output tokens, and wall time, although reasoning-token use increased. A separate 273-row run demonstrated operation at a larger test plan size while retaining unsupported environment-specific values for owner resolution.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Raj C, Bharath; Anandan, Mukund; and Gourav, Gourav, "CONSTRAINT-PRESERVING AI GENERATION OF NETWORK TEST PROCEDURES", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/12038