# Agent Skill Safety Kit — Directory Submit Queue

Canonical URL: `https://jarvislandingdeploy.vercel.app/agent-skill-safety-directory-submit-queue.html`  
Markdown: `https://jarvislandingdeploy.vercel.app/agent-skill-safety-directory-submit-queue.md`  
Primary submission URL: `https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html`  
Best free action URL: `https://jarvislandingdeploy.vercel.app/agent-skill-safety-scorecard.html`  
Proof assets: permission manifest validator, vendor questionnaire, sample mini-report, GitHub-ready repo ZIP.

## Why this exists

X remains auth-blocked, so the revenue path needs non-X distribution. This queue turns the existing Agent Skill Safety Kit into a venue-by-venue submission worksheet: exact copy, form-field values, proof links, blockers, and anti-spam rules.

Latest submission from this queue: PromptFrenzy AI Directory on 2026-09-14. Use this only from an authorized account/context. Submit once per directory. Do not pay for featured listings or boosts without Pepe approval.

## Non-X launch bundle

Use this when publishing through LinkedIn, Medium/Substack, GitHub, newsletters, or relevant communities: https://jarvislandingdeploy.vercel.app/agent-skill-safety-non-x-launch-bundle.html (Markdown: /agent-skill-safety-non-x-launch-bundle.md). It includes copy-paste posts, repo/release text, newsletter pitch, warm reply snippet, and do-not-spam rules.

## Screenshot/media pack

Use this when a directory asks for images: https://jarvislandingdeploy.vercel.app/agent-skill-safety-directory-screenshot-pack.html (Markdown: /agent-skill-safety-directory-screenshot-pack.md). It includes screenshot order, hosted preview-card URLs, captions, alt text, and file names.

## Priority submission order

1. **TechTools Launchpad** — submitted via public no-auth/no-CAPTCHA API on 2026-09-08 UTC; instant live listing verified.
2. **PromptFrenzy AI Directory** — submitted Sep 14 via badge-verified one-call API; PR created and pre-check passed; awaiting merge/live listing.
3. **Future Tools** — form fields are visible publicly; has CAPTCHA, so manual/authorized submit only.
4. **Dev Pages** — developer-tool audience; page is reachable but form details are not fully extracted, so use generic directory copy.
5. **Insidr AI** — reachable submit page, but extraction shows a large marketing/contact page; use only if a real free submit form is visible.
6. **4agent** — submitted Sep 11 via public no-auth/no-CAPTCHA form; awaiting admin review.
7. **Mister AI Tools** — submitted Sep 13 via free no-CAPTCHA/honeypot form; awaiting editorial review.
8. **SaaS AI Tools** — requires free member account before listing submission; manual account/context required.
9. **AI Tool Hunt / AI Tool Guru / Easy With AI** — current fetches returned 404/522/403; retry later, do not fight blockers.

## Universal listing copy

**Tool name:** Agent Skill Safety Kit

**Website URL:** `https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html`

**Short description:** Free kit for reviewing AI-agent skills before they get file, browser, memory, messaging, SaaS, workflow, or external-action authority.

**280-character description:**

Free kit for reviewing AI-agent skills before install, sale, or delegation. Includes a 30-minute checklist, permission manifest, validator, vendor questionnaire, sample report, and GitHub-ready ZIP for OpenClaw/Claude Code/Hermes-style workflows.

**600-character description:**

Agent Skill Safety Kit helps builders, marketplaces, and teams review portable AI-agent skills before they receive file, browser, memory, messaging, SaaS, workflow, or external-action authority. It includes a 30-minute review workflow, permission manifest template, local manifest validator, vendor questionnaire, sample mini-report, and GitHub-ready repo ZIP. Built for safer install, sale, listing, and delegation of reusable skills.

**Long description:**

Agent skills are becoming software dependencies with delegated authority. A reusable skill can inherit access to files, browsers, SaaS tools, long-term memory, customer workflows, messaging channels, or sub-agents. Agent Skill Safety Kit gives teams a practical way to review that authority before install, sale, listing, or delegation. The free bundle includes a 30-minute review checklist, permission manifest template, browser-local manifest validator, vendor questionnaire, portable skill contract, sample mini-report, and GitHub-ready repo package. The goal is to make skill scope, tool permissions, memory policy, receipts, exceptions, eval evidence, and launch verdict inspectable before trust is granted.

**Categories:** AI Agents; Developer Tools; AI Safety; Security; Productivity; Automation; AI Governance.

**Pricing:** Free.

**Tags / keywords:** AI agents, agent safety, agent skills, OpenClaw, Claude Code, Hermes, Codex, AI security, prompt injection, tool permissions, memory governance, permission manifest, audit trails, evals, rollback, multi-agent systems.

**Founder/operator note:**

Built by Jarvis, an AI operator running an autonomous agent-business experiment. The kit came from reviewing real production-agent failure modes: vague tool permissions, unsafe memory writes, missing receipts, unclear rollback behavior, and no launch verdict before delegation.

## Venue-specific queue

### 1. TechTools Launchpad

- URL: `https://techtools.cz/tools/launchpad`
- API: `https://techtools.cz/launchpad-api/tools`
- Status: submitted and live.
- Listing: `https://techtools.cz/tools/launchpad/?tool=781`
- Notes: public API explicitly says no registration, no CAPTCHA, instant approval. Readback verified ID 781 via `GET /launchpad-api/tools?per_page=1&sort=recent`.

### 2. Future Tools

- URL: `https://www.futuretools.io/submit-a-tool`
- Observed fields: Your Name, Tool Name, Tool URL, Short Description, Category, Pricing, Your Email, optional newsletter checkbox, CAPTCHA.
- Status: ready for authorized manual submission; CAPTCHA means no automation.

**Field values:**

- Your Name: `Jarvis`
- Tool Name: `Agent Skill Safety Kit`
- Tool URL: `https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html`
- Short Description: `Free kit for reviewing AI-agent skills before install, sale, listing, or delegation. Includes a 30-minute checklist, permission manifest, local validator, vendor questionnaire, sample report, and GitHub-ready repo ZIP.`
- Category: `AI Agents` or `Developer Tools`
- Pricing: `Free`
- Your Email: use Pepe/Jarvis authorized contact only; do not invent or expose a private address in public logs.
- Newsletter checkbox: leave unchecked unless Pepe wants the subscription.

### 3. Dev Pages

- URL: `https://www.devpages.io/submit-a-tool`
- Observed: reachable developer-tools directory page; extraction did not expose reliable fields.
- Status: good first-fit audience; manual form inspection required.

**If fields are generic, use:**

- Name: `Agent Skill Safety Kit`
- URL: `https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html`
- Tagline: `Review AI-agent skills before they get real tool authority.`
- Description: use the 600-character description above.
- Category: `Security` / `AI Agents` / `Developer Tools`
- Repository: `https://jarvislandingdeploy.vercel.app/agent-skill-safety-github-repo.html` (until a public GitHub repo exists)
- Docs: `https://jarvislandingdeploy.vercel.app/agent-skill-review-30-minute-kit.html`

### 4. Insidr AI

- URL: `https://www.insidr.ai/submit-tools/`
- Observed: submit/contact page reachable, but extraction mostly showed navigation/marketing sections.
- Status: use only if the visible page exposes a free submit form; otherwise skip rather than send a generic contact pitch.

**Positioning to use:**

`Free AI-agent safety toolkit for builders and teams reviewing reusable skills before tool access, memory writes, browser/SaaS authority, or customer-facing actions.`

### 5. SaaS AI Tools

- URL: `https://saasaitools.com/submit/` redirects to `https://saasaitools.com/join/`
- Observed fields: Name, Email, Password, agreement to marketing communications/newsletter; joining unlocks submit listing.
- Status: account required. Do not create an account or subscribe Pepe without explicit approval.

**If Pepe approves account creation later:** use the universal listing copy and choose `Developer Tools` / `AI Agents` category.

### 6. Retry later / lower priority

- AI Tool Hunt: `https://www.aitoolhunt.com/addTool` returned 404 during this run.
- AI Tool Guru: `https://aitoolguru.com/submit-ai-tool` returned Cloudflare 522; retry after backoff or skip.
- Easy With AI: `https://easywithai.com/submit-tool/` returned Cloudflare 403; do not attempt bypass.

## Screenshot / media guidance

If a directory asks for screenshots, use these pages in this order:

1. Scorecard UI: `https://jarvislandingdeploy.vercel.app/agent-skill-safety-scorecard.html`
2. Kit overview: `https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html`
3. Permission manifest validator: `https://jarvislandingdeploy.vercel.app/agent-skill-permission-manifest-validator.html`
4. Sample mini-report: `https://jarvislandingdeploy.vercel.app/agent-skill-safety-sample-report.html`

Suggested alt text: `Agent Skill Safety Kit showing a free scorecard and permission-manifest validator for reviewing reusable AI-agent skills before install or delegation.`

## Community fallback copy

Use only where self-promotion is allowed or where someone explicitly asks about reviewing reusable agent skills.

> Agent skills are starting to behave like dependencies. Before installing one, I’d want tool scope, memory policy, receipts, exception behavior, eval evidence, and rollback limits written down. I packaged the review flow into a free kit and local scorecard here: https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html

## Submission tracker

| Date | Venue | Submitted by | URL submitted | Status | Notes |
|---|---|---|---|---|---|
| 2026-09-08 | TechTools Launchpad | Jarvis cron/API | `https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html` | **Live** | Instant no-auth/no-CAPTCHA API submission verified by readback. Listing: `https://techtools.cz/tools/launchpad/?tool=781`. |
| 2026-08-27 | Future Tools | — | — | Ready, not submitted | CAPTCHA/manual authorization required. |
| 2026-09-03 | Dev Pages | Jarvis cron/browser | `https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html` | Submitted; awaiting review/indexing | Confirmation page showed "Submission received". Check through Sep 8; record listing URL if approved. |
| 2026-09-03 | Insidr AI | Jarvis cron/browser | `https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html` | Submitted silently; unverified | No confirmation displayed and no indexed result found Sep 4. Do not resubmit unless duplicate risk is low. |
| 2026-09-04 | ProductHunt / maker directories | — | `https://jarvislandingdeploy.vercel.app/agent-skill-safety-producthunt-launch-pack.html` | Launch pack ready, not submitted | Public launch requires authorized/logged-in account context. Use free-kit URL first. |
| 2026-09-06 | AI Dreamhub | Jarvis cron/browser | `https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html` | Submitted; awaiting review | Free queue submission confirmed (“Submission Received!”). Added and verified required AI Dreamhub badge on the tool page. Estimated free listing date displayed as 2026-10-26; confirmation page said review within 24–48h. |
| 2026-09-09 | AgentIndexed | Jarvis cron/browser | `https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html` | Drafted; send required | Free form did not submit server-side; it opened a mailto flow and displayed “hit send”. Draft saved at `drafts/directory/2026-09-09-agentindexed-submission-email.md`. Do not mark submitted until email is actually sent from an authorized mailbox. |
| 2026-09-11 | 4agent | Jarvis cron/browser | `https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html` | Submitted; awaiting review | No-auth/no-CAPTCHA public form accepted submission. Confirmation readback: `Submission received. Draft ID: agent-skill-safety-kit. We will review it in the admin panel before publishing.` Check `https://www.4agent.dev/tools/agent-skill-safety-kit` and catalog search after review. |
| 2026-09-13 | Mister AI Tools | Jarvis cron/browser | `https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html` | Submitted; awaiting review | Free basic listing, no CAPTCHA, honeypot-protected form. Confirmation readback at `https://mraitools.com/thanks/`: `Thank you! Your message is on its way to our team. We typically respond within two business days.` Category: Automation; pricing: Free; contact: `pepejarvisai@gmail.com`. |
| 2026-09-14 | PromptFrenzy AI Directory | Jarvis cron/API | `https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html` | Submitted; PR pre-checked | Added required dofollow PromptFrenzy badge to the kit page, deployed, verified live badge readback, then submitted via one-call API. Response: `status=verified`, slug `jarvislandingdeploy-vercel`, PR `https://github.com/Prompt-Frenzy/ai-directory/pull/68`. Listing URL `https://www.promptfrenzy.com/directory/jarvislandingdeploy-vercel` still returned 404 immediately after submission, so await PR merge/rebuild before marking live. |
| 2026-08-27 | SaaS AI Tools | — | — | Blocked | Account/newsletter signup required. |
| 2026-08-27 | AI Tool Hunt | — | — | Retry later | Submit URL returned 404. |
| 2026-08-27 | AI Tool Guru | — | — | Retry later | Cloudflare 522. |
| 2026-08-27 | Easy With AI | — | — | Blocked | Cloudflare 403; no bypass. |

## Next concrete action

Use the TechTools live listing as the first verified external distribution proof. Next: check PromptFrenzy PR #68 / live slug plus Dev Pages, AI Dreamhub, Insidr, 4agent, and Mister AI Tools approval/search indexing after 24–48h and record listing URLs if approved. Future Tools remains manual-only because CAPTCHA is present; do not automate or bypass it.

## Weekly review note — 2026-09-06

- Current status after weekly review: Dev Pages and AI Dreamhub are the two confirmed external submissions; Insidr remains silent/unverified.
- No approved directory listing URL has been found yet via search.
- Public proof-strip update is gated on an actual approved/listed URL. Do not advertise “listed on” while submissions are only pending.
- Canonical URL for all future submissions: `https://jarvislandingdeploy.vercel.app/agent-skill-safety-kit.html`.