jfrog/ai-agent-examples/jfrog-cli

Claude Code and agent skill by jfrog/ai-agent-examples

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Use when working with the JFrog CLI (jf command) -- uploading/downloading artifacts, running security scans, managing builds, creating release bundles, configuring the JFrog Platform from the command line, or invoking platform REST APIs via jf api. Triggers on mentions of jf command, jfrog cli, jf api, jf rt, jf audit, jf scan, jf docker, file spec, or jf config.

Install

npx vskill@latest install jfrog/ai-agent-examples/jfrog-cli

SKILL.md

JFrog CLI Skill

Installation

See Install JFrog CLI.

bash
# macOS
brew install jfrog-cli
# Linux (curl) — runs remote script; for production prefer package manager or verify script integrity
curl -fL https://install-cli.jfrog.io | sh
# Docker
docker run releases-docker.jfrog.io/jfrog/jfrog-cli jf --version

Requires JFrog CLI 2.100.0+ for `jf api` (platform REST). Other jf commands need a recent v2.x CLI.

Authentication

Follow login-flow.md to resolve the active JFrog environment. The jf CLI is required and will be installed automatically if missing. The flow supports multiple environments and persists credentials via jf config (encrypted at rest):

  1. Ensure `jf` is installed at 2.100.0+ (required for jf api).
  2. Check saved credentials -- runs jf config show to list configured servers. If multiple are saved, asks the user which environment to use.
  3. Confirm platform URL -- show URLs from jf config show; user confirms or switches server.
  4. Readiness -- jf api /artifactory/api/v1/system/readiness after confirmation.
  5. Agent-driven web login -- if no credentials exist, asks the user for their JFrog URL, drives the REST-based login flow (user clicks a link, authenticates in browser), and saves the resulting token via jf config add.

After login, credentials are saved with a server ID derived from the hostname (e.g. mycompany from mycompany.jfrog.io) so the user can switch between environments with jf config use <server-id>.

Platform REST via jf api

Prefer `jf api <path>` for JFrog Platform HTTP APIs (path only; auth from jf config). See [jf-api-patterns.md](jf-api-patterns.md) for session start (confirm URL → readiness), status/body capture, and `curl` fallback rules.

Configuration

bash
# List configurations
jf config show
# Set default server (if multiple configured)
jf config use my-server
# Add server configuration (non-interactive, using env vars)
jf config add my-server \
--url=https://$JFROG_URL \
--access-token=$JFROG_ACCESS_TOKEN \
--interactive=false
# Export/import config (for CI). Never commit the export file; add it to .gitignore.
jf config export my-server > jf-config.export
jf config import jf-config.export
Environment Variables
VariableDescription
JFROG_URLJFrog instance hostname (no https://). Used across all JFrog skills
JFROG_ACCESS_TOKENAccess token. Used across all JFrog skills and for direct REST API calls
JF_URLJFrog Platform URL (CLI-native alternative to JFROG_URL)
JF_ACCESS_TOKENAccess token (CLI-native alternative to JFROG_ACCESS_TOKEN)
JF_USER / JF_PASSWORDBasic auth credentials
JFROG_CLI_BUILD_NAMEDefault build name
JFROG_CLI_BUILD_NUMBERDefault build number
JFROG_CLI_BUILD_PROJECTDefault project key

File Specs

JSON format for specifying upload/download/search patterns:

json
{
"files": [
{
"pattern": "libs-release-local/com/example/(*)/(*).jar",
"target": "downloads/{1}/{2}.jar",
"flat": false,
"recursive": true,
"regexp": false
}
]
}

Use with: jf rt upload --spec=filespec.json, jf rt download --spec=filespec.json

Command Groups Overview

GroupPrefixDescriptionReference
Artifactoryjf rtArtifact operations, repos, buildsartifactory-commands.md
Securityjf audit, jf scanVulnerability scanning, SBOMsecurity-commands.md
Platformjf ds, jf worker, jf evdDistribution, workers, evidenceplatform-commands.md

Most Common Commands

bash
# Upload artifacts
jf rt upload "build/*.jar" libs-release-local/com/example/app/1.0/
# Download artifacts
jf rt download "libs-release-local/com/example/app/1.0/*.jar" ./local/
# Search artifacts
jf rt search "libs-release-local/com/example/**/*.jar"
# Build integration
jf rt build-add-git my-build 1
jf mvn install --build-name=my-build --build-number=1
jf rt build-publish my-build 1
# Security scanning
jf audit # Scan project dependencies
jf scan ./myapp.jar # Scan a binary
jf docker scan myapp:1.0 # Scan Docker image
# Release bundles
jf release-bundle-create my-bundle 1.0 --builds="my-build/1" --signing-key=mykey
jf release-bundle-promote my-bundle 1.0 --environment=PROD
jf release-bundle-distribute my-bundle 1.0 --site="edge-*"
# Evidence
jf evd create --build-name=my-build --build-number=1 \
--predicate=./sign.json --predicate-type=https://jfrog.com/evidence/signature/v1 \
--key="$PRIVATE_KEY"

Parallelization

When building and pushing multiple Docker images, run the builds concurrently using parallel subagents or background processes. Docker builds are independent of each other and are typically the most time-consuming step in onboarding workflows. Each docker build + jf docker push pair can run in its own subagent. Publish build info (jf rt build-publish) only after all image pushes for that build are complete.

Reference Files

Official Documentation

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/jfrog/ai-agent-examples/jfrog-cli/badge)](https://verified-skill.com/skills/jfrog/ai-agent-examples/jfrog-cli)

Something wrong with this skill or its page? Tell us.