Automate SVA environment analysis by filling formal records from verification results.
npx skills add https://github.com/XS-MLVP/UCAgent --skill sva-audit
Please help me install this Agent Skill. Skill: sva-audit Source: https://github.com/XS-MLVP/UCAgent/tree/main/ucagent/lang/zh/skills/formal/sva-audit Command: npx skills add https://github.com/XS-MLVP/UCAgent --skill sva-audit
Environment analysis in SVA verification often requires manual translation of results into a formal_records.yaml. This skill guides LLMs to fill analysis details and keep records up-to-date with automated scaffolding and incremental updates.
Run the update_analysis.py script to show or fill analysis entries and then verify the records with the environment checker.
High-intent search queries and answers about installing and using this skill.
Automating SVA verification environment analysis involves using scripts to guide LLMs in filling formal analysis records from run results. This auto-scaffolds analysis entries and incrementally synchronizes checker outputs to keep records current.
Maintaining a formal_records.yaml file for audits is best done through incremental updates via dedicated scripts. This approach auto-scaffolds entries from TRIVIALLY_TRUE and FALSE properties, ensuring structured, audit-ready records without manual translation.
Yes, you can use LLM-guided scripts like update_analysis.py to update TRIVIALLY_TRUE and FALSE properties. These scripts auto-scaffold analysis entries and incrementally synchronize checker outputs into your structured records.
Yes, a .formal_records.yaml file is required to perform structured SVA environment analysis. This file serves as the central data structure for the scripts to populate and update with TRIVIALLY_TRUE and FALSE property analysis details.
To incrementally synchronize checker outputs with SVA analysis records, run the update_analysis.py script. It applies incremental scaffolding to update existing formal records with new run results, keeping your audit data continuously current.
XS-MLVPAnalyze COI coverage data and generate SVA assertions for uncovered signals.
rakhnianskii9Coordinate AI agents to plan, execute, and sign off on QA testing cycles.
bluefishsRequire fresh verification evidence before declaring task completion.
ethan-rioManage human annotation workflows for Arize LLM observability projects.
