Authority-Bound Execution for Agentic Systems: A Formal Architecture for Provenance-Validated Admission, Revocation Fencing, and Recoverable Effects

Agentic software can propose operations affecting files, accounts, infrastructure, and other external resources. Deterministic access control is not sufficient when authorization facts originate from the agent being assessed, evidence is reused outside its intended context, or authorization becomes stale before consequential effects occur. This preprint introduces Authority-Bound Execution (ABE), a proposed reference architecture coupling owner-resolved security facts, monotonic capability attenuation, context-applicable provenance, effect-time revocation fencing, and durable outcome reconciliation. The manuscript defines a formal transition model, four conditional safety propositions, explicit counterexamples, and a falsifiable experimental evaluation protocol. It distinguishes atomic local effects from remote operations with uncertain outcomes and identifies the consistency assumptions required for authorization enforcement across system boundaries. No completed experimental benchmarks, machine-checked proofs, production validation, or independent peer review are claimed. The contribution is a theoretical systems-security framework and research program rather than an empirically demonstrated improvement over existing authorization architectures. Author: Christian SerranoAffiliation: Caivra Tech LLCORCID: https://orcid.org/0009-0009-8778-5800Manuscript version: 0.5