system
Agentic Systems
Bounded tools, orchestration, evaluation, and inspectable traces for systems that can act.
Public content search
Search the public evidence.
Find published LucyAI records without exposing drafts, private notes, archived work, or no-index material.
Publication-gated index
Search what LucyAI has actually published.
Draft, archived, no-index, and private records are excluded before this index reaches the browser.
25 published records shown.
system
AI Engineering
Retrieval, context, evaluation, and production boundaries treated as engineering decisions.
system
Software Architecture
Decisions, interfaces, constraints, and viewpoints made explicit enough to inspect and maintain.
system
AI Security
Secure development, adversarial thinking, and least-privilege authority for AI-enabled systems.
system
Research & Evaluation
Source-led inquiry and evaluation plans that separate findings, analysis, limits, and open questions.
system
Technical Creation
Original technical products and publication systems shaped around explicit constraints.
research
Evaluating agents beyond the happy path
A source-led field note on tasks, trials, graders, traces, harnesses, and failure conditions.
writing
Context is architecture, not a prompt window
An editorial synthesis of retrieval, information placement, representation, and evaluation boundaries.
research
Least-privilege tools for autonomous systems
A security synthesis on purpose-bound authority, validation, observable effects, and recovery.
workbench
LucyAI Web Platform
A verified public platform build that turns LucyAI's evidence-first publishing model into an inspectable production artifact.
workbench
Agentic Systems Field Guide
An editorial-preview concept for navigating agent architecture, tools, controls, traces, and evaluation.
workbench
AI Risk & Evaluation Atlas
A concept study for connecting risks, evaluation questions, evidence, and unresolved uncertainty.
resource
Evidence register template
A printable structure for connecting public claims to provenance, scope, review, and publication state.
resource
Agent evaluation canvas
A printable canvas for tasks, trials, graders, traces, harness conditions, failures, and review decisions.
resource
Source map for accountable AI systems
A source-led map of the official and primary references used by LucyAI's launch editorial registry.
source
Building effective agents
vendor-first-party source reviewed 2026-07-11.
source
Demystifying evals for AI agents
vendor-first-party source reviewed 2026-07-11.
source
NIST AI 600-1: Generative Artificial Intelligence Profile
government-guidance source reviewed 2026-07-11.
source
Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
primary-research source reviewed 2026-07-11.
source
Lost in the Middle: How Language Models Use Long Contexts
primary-research source reviewed 2026-07-11.
source
Introducing Contextual Retrieval
vendor-first-party source reviewed 2026-07-11.
source
NIST SP 800-218A: Secure Software Development Practices for Generative AI
government-guidance source reviewed 2026-07-11.
source
NIST AI 100-2 E2025: Adversarial Machine Learning
government-guidance source reviewed 2026-07-11.
source
OWASP Top 10 for Agentic Applications 2026
community-guidance source reviewed 2026-07-11.
source
ISO/IEC/IEEE 42010:2022
standard source reviewed 2026-07-11.