AI generates more changes. Understand what they could affect.
Setsumei maps how your code connects and prioritizes potential risks. Explore the results in a local scan brief, or let a coding agent query the analysis through MCP before planning an edit.
SETSUMEI SCAN BRIEF · DEMONSTRATION DATA
More change, same need to understand the system.
Teams already combine source reading, IDE navigation, static analysis and expert knowledge. As proposed changes multiply, assembling that context for the next question takes experienced attention.
See what deserves a closer look and why.
A finding in a widely used component may deserve earlier attention than the same pattern in isolated code. The brief brings priority, relationships and supporting source into one place to investigate.
Enlarge the example brief
Reuse the same context for the next question.
The next task does not have to begin with reconstructing every relationship.
Use the analysis to get oriented, prioritize potential issues, document architecture or plan a refactor. Engineers explore the report and local brief; compatible coding agents query dependencies, findings, source and freshness through MCP.
The core static analysis is deterministic and builds the relationships and candidate findings without an LLM or model API key. An optional AI-assisted investigation step calls a language model to examine selected candidates. Refresh after relevant changes so the context matches the code you are working on.
Understand the codebase. Test the agent's behavior.
Setsumei provides reusable codebase analysis for engineers and coding agents. RunQC separately tests AI-agent behavior through a permitted endpoint. Choose the product for the question you need to answer.
Setsumei and RunQC →RunQC exercises a permitted AI-agent endpoint across behavioral, security, reliability, and domain-specific checks, then returns the interactions and results for human review.
Explore RunQC ↗Bring one codebase question.
Discuss a local evaluation of the scan, browser brief and workflow your team needs, from understanding unfamiliar code to planning the next change.
Working with client codebases? See the services-firm use case.