Analyze code repositories read-only with confidence levels and evidence boundaries.
npx skills add https://github.com/koasis89/ite-ai-agent --skill analyze-koasis89
Please help me install this Agent Skill. Skill: analyze Source: https://github.com/koasis89/ite-ai-agent/tree/main/plugins/oh-my-codex/skills/analyze Command: npx skills add https://github.com/koasis89/ite-ai-agent --skill analyze-koasis89
This Skill solves the problem of needing a clear, confidence-laden analysis of a repository's codebase to understand its architecture, behavior, or other aspects without making any changes.
Analyze the repository behavior related to 'feature X' by executing the command '$analyze feature X'.
High-intent search queries and answers about installing and using this skill.
Perform a read-only repository analysis to synthesize insights into architecture and behavior. This deep, cross-file analysis answers complex questions about your codebase without making any changes to the source files.
Confidence-based synthesis provides ranked explanations with explicit confidence levels during code analysis. It establishes clear boundaries between explicit evidence found in the repository and logical inferences made by the analyzer.
Analyze repository behavior related to the feature by executing a targeted command like '$analyze feature X'. This performs a comprehensive, cross-file synthesis to explain system behavior with ranked confidence levels.
Yes, repository analysis deduces the impact a change would have on the system. By performing deep, read-only analysis, you understand architectural dependencies and behavioral logic without the risk of modifying the code.
No, deep repository analysis requires read access only. It explicitly avoids modifying the codebase, ensuring a safe, non-destructive environment for synthesizing architectural and behavioral insights.
Read-only analysis synthesizes insights without executing the code, meaning runtime behaviors are inferred rather than observed. Conclusions are bounded by explicit evidence versus inference, ensuring transparent confidence scoring.
riftzen-bitInstall and configure Everything Gemini Code components with verification and optional optimization.
henrique-simoesConvert research notes and transcripts into structured executive reports.
rohanbhatia27Enhances code graphs with SARIF audit findings mapped to nodes by file and line overlap.
Weihong-LiuAnalyze codebases read-only and rank evidence-based explanations with confidence levels.
