An inference-time framework for long-horizon LLM agents with explicit belief states, consistency validation, and trapping-aware recovery.
AI agent skill that converts Keil MDK/uVision projects to reproducible CMake and GNU Arm builds, with validation and bilingual build guides.
Fail-closed completion validation and path ownership for coding agents.
—
Seedance 2.0 agent skill with evidence gates, reference roles, sequence state, repair loops, validation, and documented API integration boundaries.
A secure, production-ready Python CLI wrapper for Nmap that makes powerful scans easy and repeatable. Includes 12 built-in profiles (SYN, aggressive, OS detection, vulnerability/NSE checks), interactive and non-interactive modes, strict input validation, Docker support, CI, and script-friendly flags for automation.
Evidence-driven forensic investigation: how a near-perfect A-share ML prediction signal collapsed under rigorous validation