What is session archaeology in Claude Code?

Session archaeology is the process of excavating Claude's local session logs to find the specific context, prompts, and subagent actions that led to a particular code change or architectural decision.

What are the technical requirements for using this skill?

Session Chronicle requires jq for JSON parsing, brotli for log compression, the AWS CLI for S3 sharing, and the 1Password CLI (op) for secure credential injection.

How does it handle auto-compacted sessions?

It identifies and links UUID chains across multiple sessions, allowing you to maintain a continuous thread of provenance even after Claude Code auto-compacts or starts a new session.

Why does this skill track subagent sessions separately?

Subagent sessions (like those created by Explore or Plan agents) contain critical reasoning that is often lost when only the main session is tracked. Recording both ensures complete provenance.

Does Session Chronicle commit large log files to my Git repository?

No. It follows a strict file ownership model where large artifacts are stored in gitignored directories or uploaded to S3, while only small, append-only registry files are committed to Git.

Session Chronicle

Name: Session Chronicle
Author: terrylica

byterrylica

•

開発者ツール

Excavates and documents Claude Code session logs to establish complete provenance for research, ADRs, and code contributions.

概要

Session Chronicle provides a robust framework for 'session archaeology,' allowing developers to trace the origin of code and decisions across multiple auto-compacted Claude Code sessions. By recording session UUIDs (both main and subagent), attributing actions to GitHub users, and managing research artifacts through S3 or local registries, it ensures that every architectural decision and research finding is reproducible and fully documented for future maintainers. This skill is essential for teams following ADR-driven development or those requiring high levels of traceability in AI-assisted workflows.

主な機能

Includes built-in preflight checks for jq, brotli, and AWS CLI dependencies
Supports artifact sharing via S3 for team-wide access to research history
Generates self-contained registry entries for research sessions and findings
Traces UUID chains across auto-compacted main and subagent sessions
8 GitHub stars
Provides automated GitHub attribution for all documented decisions

ユースケース

Documenting research findings with full session context for long-term reproducibility
Answering 'who created this?' by tracing code back to specific AI agent iterations
Linking Git commits to specific Claude Code session UUIDs for audit trails

概要

主な機能

Includes built-in preflight checks for jq, brotli, and AWS CLI dependencies
Supports artifact sharing via S3 for team-wide access to research history
Generates self-contained registry entries for research sessions and findings
Traces UUID chains across auto-compacted main and subagent sessions
8 GitHub stars
Provides automated GitHub attribution for all documented decisions

ユースケース

Documenting research findings with full session context for long-term reproducibility
Answering 'who created this?' by tracing code back to specific AI agent iterations
Linking Git commits to specific Claude Code session UUIDs for audit trails