I wrote this prompt/skill because of this LLM failure mode: when you ask an LLM to "write tests," it inevitably spits out feature-shaped, data-flow, happy-path assertions. It writes examples, not properties.
The solution is like this: prompt in a way that forces a formal decomposition workflow that activates more of the formal verification part of LLM "Brain".
Network Software Engineer (P2 @ Ciena) specialized in formal concurrency verification and high-performance system observability. I focus on "Architecture Over Intelligence"—building reliable agentic workflows where correctness is enforced by the system design rather than probabilistic model outputs.
Key Projects & Impact:
Clotho: Built a deterministic interleaving orchestrator that uses AST-level hijacking to verify Python codebases. It prunes $O(N!)$ state spaces into $O(2^{C})$ Mazurkiewicz equivalence classes via Source-DPOR.
Project Oracle: Architected an AI-native diagnostic suite that slashed TTR for core FIB (Forwarding Information Base) issues from 4 hours to <3 seconds by establishing a C-state-to-JSON "lingua franca" for LLM reasoning.
Knowledge Recovery: Successfully decoded complex, undocumented legacy control planes into formal specifications to enable critical 6500-to-10.X system transitions.
LTL-Driven Testing: Built a permutation-based framework using Linear Temporal Logic to auto-generate 8,000+ high-fidelity tests, identifying regressions in minutes that previously stalled teams for weeks.
I am looking for a new Senior Backend or Systems Engineering role where I can apply rigorous formal methods to modern AI-native or high-concurrency financial/distributed infrastructures.
I wrote this prompt/skill because of this LLM failure mode: when you ask an LLM to "write tests," it inevitably spits out feature-shaped, data-flow, happy-path assertions. It writes examples, not properties.
The solution is like this: prompt in a way that forces a formal decomposition workflow that activates more of the formal verification part of LLM "Brain".