Chainloop: AI coding session lineage view

Hi, we're presenting the new AI Code Session Lineage View.

It represents the entire lifecycle of AI-generated code, from the developer's prompt to production.

It uses Chainloop Trace to gather all the session material, including metadata such as the harness, the models used, token usage, and cost, as well as code attribution to either a human or the AI.

It also gathers all the context used during the session, such as linear issues, specs, skills, and the full transcript for further processing.

Chainloop collects all this metadata as signed attestations and correlates it with the developer's PRs on both GitHub and GitLab.

As part of the integration, Chainloop runs a set of automated scans on the PR diff to find any security vulnerabilities or SAST findings introduced in the code.

It also runs AI-assisted scoring to measure how well the PR aligns with the specs and the quality of the final implementation.

The lineage view closes the loop on the AI SDLC by correlating the code produced in the AI session with the final artifacts built from the merged code across the organization's CI pipelines.

Finally, it identifies the vulnerability findings that affect those artifacts.

This helps you see whether the session introduced a possible regression or vulnerability.