The /insights you wish you had.
vibe-insights reads your Claude Code transcripts across every machine and your whole history, then hands back coverage, where-was-I recall, the cache reveal, trends, the open threads worth picking back up, and a synthesized read of how you actually work. Personal by default, no telemetry, all local. The verbose /insights, over everything.
Includes1 skill
ValidatedProven on a live 195-session personal index across two machines.
Everything /insights leaves on the table.
A deterministic engine reads your session transcripts and produces one branded report — personal by default, split by source and machine, with anything you mark private kept local.
Where was I, across every machine.
Sessions, repos, and burn split by source × machine. Plus your most-recent sessions with titles, branches, and machines — ADHD-brain recall for where you left off.
The number the flat report buries.
Burn, output share, and the cache-read tokens carrying most of your input-side context. The headline burn is the small number; vibe-insights shows you the rest.
The threads worth resuming.
Open feature branches you walked away from, ranked by how likely they are to be genuinely unfinished — plus a prune list for the ones that look shipped or abandoned.
A read, not a count.
Your tool mix (the build-debug loop), delegation habit (subagent + Haiku usage), and per-session friction, satisfaction, and outcome — rolled up by deterministic charts, then a synthesized narrative read grounded in the numbers.
More in the same report.
Every run produces these too — no extra commands.Burn per day, recent-vs-baseline acceleration.
Which box is the workhorse, side by side.
What you edit, tool errors, parallel sessions.
Your logged architectural decisions — file or dashboard.
Keep any source or repo local-only — never synced, never merged.
Deterministic engine, LLM only where it earns it.
The engine is deterministic and offline. It never calls an LLM or MCP — it reads local JSONL transcripts and a few cache files. The narrative read, the decisions overlay, and the per-session tags come from the /vibe-insights skill, where the LLM lives, and are handed to the engine as files. That keeps the analytics reproducible and the tool marketplace-portable.
Personal by default. Every ~/.claude* source on your machine is included and synced-eligible — no config to get started. Cross-machine works through tiny per-machine indexes you sync (Syncthing or anything that moves a folder); the report merges every machine. When you need to wall off an employer or client, mark a source or repo private — it stays in your own report but never leaves its origin box. Runs fully standalone; the optional cc-logs package keeps token numbers aligned with sibling tools like Sanduhr. Decisions are MCP-agnostic: a file, or your dashboard.
Two channels.
Stable marketplace
Tagged releases via the Vibe Plugins marketplace. The Python engine installs from the repo; the optional cc-logs extra keeps token numbers aligned with sibling tools.
/plugin marketplace add estevanhernandez-stack-ed/vibe-plugins /plugin install vibe-insights@vibe-plugins
Canary bleeding edge
Tracks this repo's main branch — faster feedback, occasional breakage.
/plugin install vibe-insights@estevanhernandez-stack-ed/vibe-insights
Personal by default. Reads every ~/.claude* source it finds; mark any source or repo local-only via the advanced block.
One plugin in a family.
Vibe Plugins are a coordinated family — installed independently, composed when present.
