agentleFS
Badhon-00Badhon-00/MDCP/skills/creator-security-automation/SKILL.md

creator-security-automation

Skill2 starsChanged 5 days agoMITView on GitHub

One cross-agent workflow for defensive cybersecurity, website protection, E2E quality assurance, and complete rights-safe YouTube content automation.

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 11 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈11

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use this as the compact entrypoint for creator workflows and defensive application security.

    line 8
  2. Distinctive:

    Test only owned systems or targets covered by explicit written authorization and a defined scope.

    line 20
  3. Distinctive:

    Default to passive inspection, source review, safe test fixtures and staging environments.

    line 21
  4. Distinctive:

    No credential theft, phishing, persistence, malware, stealth, destructive exploitation, unauthorized scanning, data exfiltration or …

    line 22
  5. Distinctive:

    Do not run SQL injection/XSS payloads against third-party production systems.

    line 23
  6. Distinctive:

    Never evade YouTube Content ID, copyright claims, licensing terms or platform enforcement.

    line 24
  7. Distinctive:

    Do not publish, change visibility, delete, comment, or mutate playlists without specific approval.

    line 25
  8. Distinctive:

    Use prepared statements and safe ORM parameter APIs.

    line 47
  9. Distinctive:

    Treat escaping and input validation as defense-in-depth, not substitutes for parameterization.

    line 49
  10. Distinctive:

    Run QA: aspect ratio, codec, loudness, captions, spelling, thumbnail readability and rights checklist.

    line 67
  11. Distinctive:

    Add to playlists, monitor analytics and produce a concise performance report.

    line 70
  12. Distinctive:

    Keep this file as the single umbrella entrypoint.

    line 79
  13. Very specific:

    Use one mdcp tool and one operation per call where the host supports it.

    line 80

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

MDCP Unified Creator + Security Automation

Use this as the compact entrypoint for creator workflows and defensive application security. Route each request to only the relevant section. The unified MCP tool is named mdcp; YouTube's existing dedicated youtube tool remains available as a compatibility adapter.

Fast router

  • Website security, vulnerability discovery, hardening → Defensive security
  • SQL injection, XSS, auth/session, API safety → Application protection
  • Browser journeys, regression, E2E, accessibility → Testing
  • YouTube channel growth, ideas, scripts, calendar → Creator strategy
  • Thumbnails, captions, editing, Shorts, audio → Production
  • Upload, scheduling, SEO, playlists, analytics → Publishing
  • Copyright claims, music/footage licensing, reused content → Rights management

Safety boundaries

  • Test only owned systems or targets covered by explicit written authorization and a defined scope.
  • Default to passive inspection, source review, safe test fixtures and staging environments.
  • No credential theft, phishing, persistence, malware, stealth, destructive exploitation, unauthorized scanning, data exfiltration or bypassing access controls.
  • Do not run SQL injection/XSS payloads against third-party production systems. Validate defenses in local labs or authorized staging with synthetic data.
  • Never evade YouTube Content ID, copyright claims, licensing terms or platform enforcement. Do not disguise copied work as original.
  • Do not publish, change visibility, delete, comment, or mutate playlists without specific approval. Uploads default to private.

Defensive security

Authorized discovery

  1. Confirm asset owner, written authorization, in-scope domains/IPs, exclusions, test window, rate limits and emergency contact.
  2. Inventory public routes, APIs, authentication boundaries, dependencies and data classes without collecting unnecessary personal data.
  3. Begin with passive TLS/HTTP header review, configuration review, dependency advisories, secret scanning and source-code review.
  4. Classify findings by severity, exploitability, impact, evidence, affected component and remediation.
  5. Retest fixes in staging; record residual risk and evidence.

Vulnerability coverage

  • Injection: SQL/NoSQL/command/template injection; use parameterized queries and safe APIs.
  • Browser: XSS, CSRF, clickjacking, CORS misconfiguration and unsafe DOM sinks.
  • Identity: weak authentication, MFA gaps, session fixation, insecure cookies, password reset and account enumeration.
  • Authorization: IDOR/BOLA, role escalation, tenant isolation and server-side access checks.
  • API: schema validation, rate limiting, mass assignment, excessive data exposure and object-level authorization.
  • Infrastructure: TLS, security headers, exposed admin surfaces, secrets, dependency/container issues and least privilege.
  • Data: encryption in transit/at rest, retention, backups, logging redaction and privacy controls.

SQL injection protection

  • Bind every user-controlled value; never concatenate user input into SQL.
  • Use prepared statements and safe ORM parameter APIs; audit raw-query escape hatches.
  • Apply least-privilege database roles, allowlisted dynamic identifiers, safe error handling and regression tests.
  • Treat escaping and input validation as defense-in-depth, not substitutes for parameterization.
  • Review source statically first; verify with synthetic fixtures in a controlled test database.

Website protection checklist

HTTPS/TLS; HSTS; CSP; frame protection; nosniff; Referrer-Policy; Permissions-Policy; secure/HttpOnly/SameSite cookies; CSRF protection; server-side authorization; rate limits; dependency patching; secret management; backups; audit logging; alerting; secure error handling; WAF/CDN rules where appropriate.

E2E and quality testing

Create a test matrix for critical journeys, roles, negative cases, API contracts, accessibility, mobile layouts, browser compatibility, network failures, retries and data cleanup. Prefer Playwright for browser automation. Use staging, synthetic accounts and isolated test data. Capture failure traces with secrets and personal data redacted. Include regression tests for every fixed vulnerability.

YouTube creator automation

Full workflow

  1. Channel audit and audience/content-pillar strategy.
  2. Research topics and viewer intent; build a publishing calendar.
  3. Generate brief, hook, script, storyboard, shot list and narration.
  4. Prepare rights-cleared footage, images, music, fonts and voice assets.
  5. Generate thumbnail concepts and produce a final image using an image tool.
  6. Edit video/Shorts: assemble timeline, crop/reframe, captions, audio mix, transitions, color and export.
  7. Run QA: aspect ratio, codec, loudness, captions, spelling, thumbnail readability and rights checklist.
  8. Generate title, description, chapters, tags and pinned-comment draft.
  9. Preview; request approval; upload privately or schedule only after explicit approval.
  10. Add to playlists, monitor analytics and produce a concise performance report.

Automation capabilities and integration boundaries

Use the YouTube Data API for authorized channel metadata, uploads and playlists; YouTube Analytics API for owner-only analytics. Use FFmpeg or a selected editor for deterministic media transforms; use image-generation tools for thumbnails; use speech/transcription tools for narration and captions when configured. Long-running rendering/upload tasks need a durable job queue, worker, object storage and resumable/idempotent jobs. The skill describes orchestration; it does not itself provision those cloud services or grant API permissions.

Rights and copyright management

Maintain an asset ledger containing source, creator, license/permission, allowed platforms, territories, term, attribution and proof. Use original or properly licensed music, footage, images, fonts and voice. Check for third-party claims through official platform workflows; document licenses and dispute only when there is a good-faith legal basis. Prefer original commentary, analysis, education or transformation with meaningful new expression, but do not assume that any label or edit automatically makes a use fair or lawful. Never promise copyright-free status or guaranteed monetization.

Token-efficient execution

  • Keep this file as the single umbrella entrypoint; load only the matching workflow.
  • Use one mdcp tool and one operation per call where the host supports it.
  • Return bounded findings, top risks and concise evidence; do not dump raw logs or full analytics.
  • Cache stable metadata; fetch volatile metrics only when needed.
  • Delegate deterministic rendering, test execution and report formatting to configured tools.
  • Inference tokens cannot be reduced to zero; savings depend on host, model, tool schemas and task size.

Expected output

Return: Scope, Actions, Findings/artifacts, Approvals needed, Verification status, Next step. Clearly distinguish a plan from a test actually executed.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.