Pp Chrome History

<!-- GENERATED FILE — DO NOT EDIT.
     This file is a verbatim mirror of library/productivity/chrome-history/SKILL.md,
     regenerated post-merge by tools/generate-skills/. Hand-edits here are
     silently overwritten on the next regen. Edit the library/ source instead.
     See the repository agent guide, section "Generated artifacts: registry.json, cli-skills/". -->

pp-chrome-history

Prerequisites: Install the CLI

This skill drives the chrome-history-pp-cli binary. You must verify the CLI is installed before invoking any command from this skill. If it is missing, install it first:

1. Install via the Printing Press installer. It defaults binaries to $HOME/.local/bin on macOS/Linux and %LOCALAPPDATA%\Programs\PrintingPress\bin on Windows:

   npx -y @mvanhorn/printing-press-library install chrome-history --cli-only

2. Verify: chrome-history-pp-cli --version
3. Ensure the reported install directory is on $PATH for the agent/runtime that will invoke this skill.

If the npx install fails (no Node, offline, etc.), fall back to a direct Go install (requires Go 1.26.6 or newer). This installs into $GOPATH/bin (default $HOME/go/bin), so add that directory to $PATH instead:

go install github.com/mvanhorn/printing-press-library/library/productivity/chrome-history/cmd/chrome-history-pp-cli@latest

If --version reports "command not found" after install, the runtime cannot see the binary directory on $PATH. Do not proceed with skill commands until verification succeeds.

chrome-history-pp-cli reads your local Chrome history SQLite database, snapshots it to ~/.cache/chrome-history/, builds an offline full-text index, and answers questions about your browsing. Read-only, zero network — nothing leaves the machine. Every query command supports --json and --select, and the same surface is exposed as MCP tools via chrome-history-pp-cli mcp — the query tools are read-only; only sync and archive_enable/archive_disable mutate local state (on-device, never the network). An optional accumulating archive (off by default) can retain history that Chrome later prunes or the user clears — see Archive below.

When to use

Historical Chrome activity from the History database: "find that page I saw," "what have I been researching," "what topics has Chrome grouped my browsing into," "how much time on X," "what did I download," "have I visited this site before."

When NOT to use (anti-triggers)

• Non-macOS — macOS only. The Chrome DB path and Full-Disk-Access model are macOS-specific; Linux/Windows are not yet supported.
• Live/open tabs — these are NOT in the History SQLite DB (Chrome keeps them in binary session files). This tool is history, not current tabs.
• Safari history — a separate CLI (Safari stores history differently; chrome-history-pp-cli only reads Chrome).
• Bookmarks, passwords, autofill, cookies — out of scope; this reads the History DB only.

Categorization (for agents)

The domains static category map is coarse; journeys exposes Chrome's own (noisy) clusters. For real topic categorization, read the --json titles/URLs and infer topics yourself — agent inference beats both the static map and Chrome's clusters (especially for clustering history into a personal vault).

Setup

Ensure chrome-history-pp-cli is on PATH (e.g. go install it, or symlink the built binary into ~/go/bin). Then snapshot Chrome's history once (re-run anytime to refresh — Chrome locks its DB, so the tool copies it safely even while Chrome is open):

chrome-history-pp-cli sync      # build/refresh the local snapshot
chrome-history-pp-cli doctor    # health check; warns if Chrome's DB schema drifts from the tested version

If doctor reports the Chrome DB is unreadable, grant your terminal Full Disk Access (System Settings → Privacy & Security → Full Disk Access).

You do not need to sync before answering a question — query commands read the snapshot that already exists. Only sync to pull in newer browsing (or the first time, if no snapshot exists yet).