Skip to content

Add a separate AI agents overview #86

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
ehl-jf merged 1 commit into from
Nov 2, 2026
Merged

Add a separate AI agents overview #86

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
23 changes: 23 additions & 0 deletions ai-agents.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,23 @@
# Integrate JFrog Workers with AI Agents

Extend your AI agents with custom, deterministic actions powered by JFrog Workers. Combine the flexibility of AI-driven decisions with predictable, repeatable execution to ensure critical software supply chain operations follow defined logic and policies.

Explore ready-to-use integration samples and practical use cases to turn agent decisions into reliable actions across your DevSecOps workflows.

Each sample folder includes `worker.ts`, `manifest.json`, and a README with the secrets, properties, and setup steps.

## After Download

These Workers run after Artifactory has served the artifact. They can record or notify. They do not block the download.

* [NanoClaw Agent Vault audit log](samples/artifactory/AFTER_DOWNLOAD/nanoclaw-agent-vault-audit-log) records a download that NanoClaw's Agent Vault resolved through Artifactory and posts that record to an audit endpoint.
* [NanoClaw Slack agent trigger](samples/artifactory/AFTER_DOWNLOAD/nanoclaw-slack-agent-trigger) posts the download to Slack so a live NanoClaw agent ingests the message and can act on it.
* [NemoClaw skill install audit](samples/artifactory/AFTER_DOWNLOAD/nemoclaw-skill-install-audit) records the skill slug, version, registry identity, and Xray status for a skill archive downloaded into a NemoClaw sandbox, then posts that record to an audit endpoint.
* [Claude download risk advisor](samples/artifactory/AFTER_DOWNLOAD/ai-download-risk-advisor) asks Claude for a short risk summary when the artifact's Xray scan has security issues or policy violations, then posts that summary to Slack. Requires Xray.
* [OpenAI download risk advisor](samples/artifactory/AFTER_DOWNLOAD/openai-download-risk-advisor) uses the same Xray lookup and Slack advisory, and calls the OpenAI Chat Completions API. Requires Xray.
* [Copilot dependency advisory](samples/artifactory/AFTER_DOWNLOAD/copilot-dependency-advisory) runs only for an artifact that has the `copilot.advisor.repo` and `copilot.advisor.repoBaseBranch` properties. On the first download, it opens a GitHub issue and assigns that issue to the GitHub Copilot coding agent.

## Before Download

* [Copilot allowlist gate](samples/artifactory/BEFORE_DOWNLOAD/copilot-allowlist-gate) allows a download when `<repoKey>/<path>` matches a glob in an allowlist file stored in GitHub. By default (`failMode=WARN`), the download proceeds with a warning if the allowlist can't be read or the Worker is not configured. Set `failMode` to `STOP` to block in that case. A folder request proceeds without an allowlist check. A blocked artifact gets a GitHub issue assigned to the GitHub Copilot coding agent.
* [Claude blocked download explainer](samples/artifactory/BEFORE_DOWNLOAD/ai-blocked-download-explainer) blocks a download when the artifact's Xray scan has more critical security issues than `maxCriticalIssues` (default `0`). Claude writes the explanation only. If Xray is unavailable, the artifact has no scan result, or the Xray search fails, the download proceeds with a warning.