Quickstart
cd your-repo
aiscope .
That's it. aiscope will scan your repo for every supported AI memory file, extract rules, find conflicts, and drop you into an interactive TUI.
What happens under the hood
When you run aiscope ., it:
- Discovers every memory file across Cursor, Claude Code, and GitHub Copilot — see Tools and subsystems.
- Parses YAML frontmatter (
applyTo,globs,tools:, …). - Extracts typed assertions (e.g. "prefer snake_case for variables").
- Reasons about pairs — checks scope overlap, polarity, severity.
- Renders the result via your chosen output mode.
It never makes a network request. It never reads outside the repo unless you
pass --user (see Privacy guard).
Output modes
| Flag | Use it for |
|---|---|
| (default) | Interactive ratatui TUI |
--text | Plain text — pipe-friendly |
--diag | Compiler-style diagnostics (miette) |
--json | Machine-readable — for scripts and dashboards |
--card out.png | 1280×720 PNG summary — drop into a tweet or PR |
Subcommands
| Command | What it does |
|---|---|
aiscope | Scan + render (default scan subcommand) |
aiscope check | Scan + exit non-zero if HIGH conflicts found (CI gate) |
aiscope watch | Re-scan on file change |
Reasoning modes
| Flag | Behavior |
|---|---|
| (default) | Uniform — every cross-source pair is candidate. Max recall. |
--specific | Specific — uses the subsystem matrix to silence false alarms. |
See Reasoning modes for the matrix.
Next
- Your first scan — walk through real output
- CLI reference — every flag, every option