Directory Persistence Audit logo

Directory Persistence Audit

Organization
zero-lab-ai
directory-persistence-audit

Audits supplied sanitized directory-service governance evidence for durable control and persistence exposure without contacting live systems. Use when reviewing defensive directory control, baseline, monitoring, or exception evidence.

Overview

Publisherzero-lab-ai
Repositoryskills
Skill namedirectory-persistence-audit
Stars
232
Forks
14
Bundled files
2
LicenseApache-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 2 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by zero-lab-ai on GitHub. Read the source before you install it.

Installation

Install the Directory Persistence Audit AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/zero-lab-ai/skills/tree/main/plugins/identity-security/skills/directory-persistence-audit
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/zero-lab-ai/skills.git /tmp/skills
mkdir -p .claude/skills
cp -r /tmp/skills/plugins/identity-security/skills/directory-persistence-audit .claude/skills/directory-persistence-audit
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Directory Persistence Audit in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Directory Persistence Audit on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Directory Persistence Audit is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Directory persistence audit

Use this skill only for passive review of an already-supplied, sanitized evidence package. It is not discovery, validation, exploitation, or authorization.

When to Use

  • Reviewing defensive directory control, baseline, monitoring, or exception evidence.
  • Use only within explicit authorization and the stated rules of engagement.

When NOT to Use

  • Do not treat this skill as permission to access, test, or change a target.
  • Do not convert missing evidence into a clean result or a stronger claim.

Boundary

Never query, contact, authenticate to, or enumerate a live directory or service. Never provide commands, scripts, LDAP or API syntax, tool instructions, object changes, persistence creation, attack paths, edges, ranking, or exploitation.

Reject distinguished names, security or object identifiers, user or service principal values, object paths, names, hosts, domains, addresses, URLs, credentials, tickets, certificates, secrets, raw access-control material, and configuration bodies. Accept only opaque IDs and digests. Diagnostics stay generic and never quote rejected content.

Workflow

  1. Accept only supplied-passive, sanitized evidence and opaque references.
  2. Record the engagement cutoff, closed policy, snapshot receipt, and opaque scope and directory references.
  3. Inventory only supplied abstract subjects. Do not discover or name objects.
  4. Create one coverage cell for every subject and each required persistence- control family.
  5. Keep control relationship, baseline, durability, protection, monitoring, freshness, and result as independent tracks.
  6. Bind every evidence item to the exact subject, snapshot and digest, scope, directory, policy and digest, control and fingerprint, and upstream receipts.
  7. Require supplied-passive mode, sanitization, eligibility, strict evidence window, cutoff compliance, and one retained custody receipt.
  8. Keep policy, relationship, baseline, durability, protection, monitoring configuration, monitoring health, result, and upstream receipt roles separate.
  9. Treat a nominal relationship as neither effective protection nor healthy monitoring. Policy assignment is neither application nor enforcement.
  10. Accept imported Kerberos or certificate-service conclusions only through exact opaque receipts bound to the same snapshot and cell.
  11. Use not-applicable only with a closed reason and exact current policy evidence. Keep that cell in the coverage denominator.
  12. Create one immutable open finding for each exact current observed-exposure. This ledger cannot close findings.
  13. Derive handoff state from exact finding, evidence, owner, and chronology. Handoffs transfer defensive work and grant no authority.
  14. Resolve structural safety failures and semantic contradictions before relying on coverage or summaries.

Interpretation rules

  • complete-no-observed-exposure means every applicable cell is current and no exposure was observed. Policy-backed NA cells do not create unknowns.
  • complete-with-exposure means at least one current observed exposure has an immutable open finding.
  • material-unknowns means a material track is unknown or indeterminate.
  • stale means historical evidence cannot support current claims.
  • incomplete means the exact coverage matrix is incomplete.

Report

Return the aggregate, track counts, open finding IDs, defensive handoffs, and generic conflict codes. Never reconstruct real directory values or operational steps from opaque evidence.

Rationalizations to reject

  • “A read-only directory query is passive.” It still contacts a live system and is outside this skill.
  • “The object name helps the reviewer.” Real names and paths create exposure; preserve only opaque references.
  • “Membership proves durable control.” A relationship is not proof of durability, protection, or effective application.
  • “Policy assignment proves enforcement.” Require separate exact evidence.
  • “Monitoring is configured, so absence is meaningful.” Configuration is not health or detection evidence.
  • “The upstream audit already passed.” Imported claims require the exact current opaque receipt and matching bindings.
  • “The label did not change.” Snapshot, policy, control, or receipt drift invalidates historical claims.
  • “We can close the finding here.” Findings stay open in this ledger.
  • “The handoff approves remediation.” Handoffs transfer defensive work only.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Directory Persistence Audit AI skill do?

Audits supplied sanitized directory-service governance evidence for durable control and persistence exposure without contacting live systems. Use when reviewing defensive directory control, baseline, monitoring, or exception evidence.

Why use Directory Persistence Audit on TypingMind?

Because you install it once and use it with any model. Directory Persistence Audit is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Directory Persistence Audit in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/zero-lab-ai/skills/tree/main/plugins/identity-security/skills/directory-persistence-audit. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Directory Persistence Audit?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Directory Persistence Audit?

As many as you like. As long as a model supports skills, you can use Directory Persistence Audit with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Directory Persistence Audit AI skill free?

Yes. It is published on GitHub by zero-lab-ai under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇