session-chronicle_skill

This skill excavates Claude Code session logs to capture complete provenance across main and subagent sessions for reproducible research.

TypeScript

14

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill terrylica/cc-skills --skill session-chronicle

SKILL.md24.7 KB

Overview

This skill provides session log provenance tracking for Claude Code workflows, capturing complete session UUID chains and creating self-contained registry entries for research findings, ADR decisions, and code contributions. It centralizes session archaeology, ensures GitHub attribution, and supports optional artifact sharing to S3 for team access.

How this skill works

The skill scans the local Claude session store for main and subagent .jsonl session files, indexes each session (entries, timestamps, type), and builds a session_contexts array that records every session UUID. It guides the user through prompts to confirm attribution, session scope, outputs, and ADR links, then writes append-only registry and iteration records, optionally compresses archives, and can upload artifacts to S3.

When to use it

You need to answer "who created this?" or trace the origin of a finding or code change
Documenting a research session or ADR with reproducible provenance
Including session UUID references in a git commit or registry entry
Tracing edits across auto-compacted main and subagent sessions
Creating a self-contained registry entry for future maintainers

Best practices

Always record ALL sessions (main + subagent) so entries are self-contained
Require GitHub username attribution for each registry entry
Keep findings/registry.jsonl and sessions/<id>/iterations.jsonl committed; place large artifacts in gitignored outputs/ and S3
Run the preflight checks to ensure jq, brotli, aws CLI, and 1Password CLI are installed
Preview and confirm the session_contexts array before writing to the registry

Example use cases

Create a registry.jsonl entry that lists every session UUID involved in a research effort
Attach session UUIDs in a structured git commit message when merging a PR tied to research
Trace a specific edit by following a UUID chain across multiple session files to find the original author
Compress session chains with Brotli and upload archived .jsonl.br to S3 for team retrieval
Link a research session registry entry to an existing ADR or create a new ADR from a finding

FAQ

Subagent sessions often contain exploration, planning, or specialized-agent context that explains decisions; omitting them forces future maintainers to re-run archaeology.

What gets committed to git versus uploaded to S3?

Only small, append-only metadata under findings/ (registry.jsonl and iterations.jsonl) is committed. Large artifacts go to outputs/ (gitignored) and can be uploaded to S3 for team sharing.