Manual distribution media kit

Screenshot pack for submitting the Agent Skill Safety Kit.

A clean, directory-ready set of screenshot targets, preview cards, captions, alt text, and file names so a manual publisher can submit without hunting through the site.

Open Markdown pack Free kit Scorecard Manifest validator

Recommended screenshot order

1. Free kit overview

agent-skill-safety-kit.html

Caption: Free kit for reviewing AI-agent skills before install, sale, listing, or delegation.

2. Browser-local scorecard

agent-skill-safety-scorecard.html

Caption: 100-point local scorecard for identity, tool authority, memory policy, receipts, rollback, evals, and portability.

3. Permission manifest validator

agent-skill-permission-manifest-validator.html

Caption: Paste a permission manifest locally and get failed checks, risky-authority summary, and a copyable validation report.

4. Sample mini-report

agent-skill-safety-sample-report.html

Caption: Example paid-review deliverable showing score, top risks, boundaries, receipts, evals, and launch verdict.

5. Vendor questionnaire

agent-skill-safety-vendor-questionnaire.html

Caption: 12 procurement/RFP questions for vendors selling or listing AI-agent skills.

Main listing URL

https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html

Use the free kit as the default directory URL; it is less salesy than the paid review page.

Existing preview cards

Use these hosted SVGs when a directory asks for square/landscape media instead of browser screenshots.

Copy-ready captions

Short caption:
Agent skills are becoming software packages. This free kit helps teams review tool authority, memory policy, receipts, rollback, eval evidence, and safe failure before install, sale, listing, or delegation.

Technical caption:
Agent Skill Safety Kit includes a 30-minute review workflow, scorecard, permission manifest template, JSON Schema, local validator, vendor questionnaire, sample mini-report, and GitHub-ready ZIP for OpenClaw/Claude Code/Hermes-style skills.

Buyer caption:
Before giving a reusable AI-agent skill access to files, browsers, SaaS, memory, messages, or workflows, ask what it can touch, what it can write, what it logs, how it fails, and how you roll it back. This kit makes those checks explicit.

File naming guidance

agent-skill-safety-kit-overview.png
agent-skill-safety-scorecard.png
agent-skill-permission-manifest-validator.png
agent-skill-safety-sample-report.png
agent-skill-safety-vendor-questionnaire.png

No submission was made by this page. Use it from an authorized browser/account only. Do not pay for boosts or featured placement without approval.