Skip to content

Repository navigation

[patch] AX-2453 - add Local Shadow AI scan section to README #101

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
yanivt-jfrog merged 3 commits into from
Nov 2, 2026
Merged

[patch] AX-2453 - add Local Shadow AI scan section to README #101

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
a81c664
[patch] docs: add Local Shadow AI scan section to README (AX-2453)
yanivt-jfrog Nov 1, 2026
f73d95c
[patch] docs: Local Shadow AI scan is on by default (AX-2453)
yanivt-jfrog Nov 2, 2026
4f4f0ee
AX-2308 - Fix rami comment
yanivt-jfrog Nov 2, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
36 changes: 36 additions & 0 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -147,6 +147,42 @@ The hook soft-fails (never breaks the session): missing project key, Agent Guard

---

## Local Shadow AI scan

The plugin can report which AI assets (skills, plugins, MCP servers) are installed
on a developer's workstation, so the organization can see what its developers'
Claude Code environments run and spot unapproved or risky assets.

**Who controls it.** The scan is off by default. It runs only when your JFrog Platform
administrator has enabled it for your account. The plugin checks this on every new session
(`GET /aisec/api/v1/local-shadow/scanning-enabled`). There is no per-user setting
in the plugin: like other endpoint security controls, it is managed by the organization.

**When it runs.** Once per new Claude Code session (not on resume, `/clear` or
compaction), in the background, and only if you are logged in to the JFrog CLI with an
access token over HTTPS.

**What it sends.** One report per session to your organization's own JFrog Platform
(`<your-platform>/aisec/api/v1/local-shadow/reports`). No scan data is sent to JFrog or
any third party. The report contains:

- A random workstation ID, generated once and stored in `~/.jfrog/shadow/workstation-id`
- The Claude Code session ID, scan time and duration, plugin and scanner versions
- For each skill: a content fingerprint, name and description
- For each plugin: name, version, marketplace, source URL and git commit
- For each MCP server: its URL (remote) or package name, version and registry (local)
- The file path where each asset was found

**What it scans.** `~/.claude`, `~/.claude.json`, and the `.claude/` folder and
`.mcp.json` of each project listed in `~/.claude.json`. It reads these config files and
sends only the fields listed above; it does not send source code, prompts, conversations
or credentials.

**Logs.** `~/.jfrog/shadow/local-shadow-hook.log` (whether the scan started and why not)
and `~/.jfrog/shadow/local-shadow.log` (scan results).

---

## Agent Package Resolution (Preview)

> **Preview Notice:** This feature is in preview and licensed under the Apache License 2.0. For clarity: This software is provided "as-is" without warranty of any kind, and without support obligations or service level commitments. Behavior, APIs, conventions, and structure may change without notice between releases. JFrog makes no guarantees of backward compatibility during the preview release cycle. Use in production environments is at your own risk.
Expand Down
Loading