Lovable Audit Stinger logo

Lovable Audit Stinger

Organization
legioncodeinc
lovable-audit-stinger

Audit Lovable-built Supabase apps: RLS denials, role-differential reads, function gating, auth posture, bundle-forensics key/endpoint extraction, findings triage, reporting.

Overview

Publisherlegioncodeinc
Repositoryvibe-coding-tools
Skill namelovable-audit-stinger
Stars
94
Forks
44
Bundled files
10
LicenseAGPL-3.0-or-later
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.

  • 10 bundled files

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

  • Open source

    Published by legioncodeinc on GitHub. Read the source before you install it.

Installation

Install the Lovable Audit Stinger 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/legioncodeinc/vibe-coding-tools/tree/main/plugins/code-audit-tools/skills/lovable-audit-stinger
  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/legioncodeinc/vibe-coding-tools.git /tmp/vibe-coding-tools
mkdir -p .claude/skills
cp -r /tmp/vibe-coding-tools/plugins/code-audit-tools/skills/lovable-audit-stinger .claude/skills/lovable-audit-stinger
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Lovable Audit Stinger 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 Lovable Audit Stinger 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 Lovable Audit Stinger 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.

Lovable Audit Stinger

This stinger audits Lovable-built applications on Supabase backends: Postgres/RLS posture, Edge Function gating, the authentication plane's configuration, and what the shipped client bundle exposes. It produces findings ledgers, control-effectiveness attestations, and known-unknowns — reports, never patches. Every factual claim it makes traces to its own research archive under references/research/; if a claim isn't in the archive, it isn't a fact yet.

When to use

  • Auditing a Lovable-built web app or its Supabase project (RLS, functions, auth, storage surface)
  • Classifying observed denials — embedded 42501-class codes, PGRST2xx, function gating — into verified controls vs findings
  • Extracting embedded keys, endpoints, and third-party hosts from a shipped JS bundle as audit evidence
  • Triaging audit observations into the operator's tier taxonomy and shaping them into a report

When not to use

  • Generic web-app security review with no Supabase/Lovable stack — route to security-stinger instead
  • Fixing what an audit finds — this stinger reports; remediation is a separate, separately-authorized engagement
  • Harness-format or Hive-component questions — queen-wasp-stinger owns creation, pest-controller-suit owns routing

Procedure

  1. Confirm the audit surface splits into its four layers — auth, data, functions, client-side exposure — and confirm authority over the target before any probe.
  2. Run bundle forensics first (guides/bundle-forensics.md): keys, endpoints, table/RPC inventories out of the shipped JS, archived with documented offsets. The inventories become the named inputs to everything downstream — nothing is guessed.
  3. Run auth-posture readout (guides/auth-posture.md): settings, confirmation gating, rate-limit shapes, token anatomy. Non-mutating or fully-disclosed synthetic probes only.
  4. Run differential visibility (guides/differential-visibility.md) per resource: the anon-vs-authenticated delta is the policy signal; reads only.
  5. Run write-denial discrimination (guides/write-denial-discrimination.md) where write-path controls must be proven: minimal labeled writes, embedded-code classification, mandatory cleanup and post-cleanup re-read.
  6. Run RPC & function-surface enumeration (guides/rpc-function-enumeration.md) on the client-named inventory: gates, role deltas, availability failures isolated from configuration.
  7. Close with audit reporting (guides/audit-reporting.md): ledger hygiene, tier assignment per the operator taxonomy, unknowns stated, hygiene section attested.

References map

  • references/probe-tables.md — load when: classifying any observed denial, shape, or gating response; the embedded-code table is authoritative over HTTP status.
  • references/worked-examples.md — load when: a probe sequence needs its validated shape before execution; six worked sequences with observed outcomes.
  • references/reporting.md — load when: shaping findings into the report; ledger schema, tier mapping, skeleton, grounding line.
  • references/research/distillation.md — load when: a domain claim needs verification or a dispute needs settling; the cited distillation of this component's own archive.
  • Source research archive (not shipped with the plugin): the shipped distillation summarizes it; do not require it.
  • guides/ (six verb guides) — load when: executing a specific audit verb; each states preconditions, procedure, validated expected shapes, anti-patterns.

Related drones and stingers

  • security-stinger — general security audit passes; owns non-Supabase surface and the Ship Gate's first gate.
  • auth-stinger — authentication-flow deep work; this stinger stops at posture readout and defers flow-level analysis.
  • pest-controller-suit — roster and routing; owns dispatch decisions and hosts this pair's registration.
  • lovable-audit-wasp-drone — the paired worker agent; the orchestrator delegates sustained audit waves to it rather than running them inline.

Critical Directive

  • You must read all files and context contained within your skill.
  • In the event your core knowledge does not provide sufficient guidance you must make every attempt to search the internet, related knowledge base documentation files, and other available resources to supplement your knowledge prior to proceeding with your task.
  • Additional related skills can be found here:

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 Lovable Audit Stinger AI skill do?

Audit Lovable-built Supabase apps: RLS denials, role-differential reads, function gating, auth posture, bundle-forensics key/endpoint extraction, findings triage, reporting.

Why use Lovable Audit Stinger on TypingMind?

Because you install it once and use it with any model. Lovable Audit Stinger 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 Lovable Audit Stinger in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/legioncodeinc/vibe-coding-tools/tree/main/plugins/code-audit-tools/skills/lovable-audit-stinger. 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 Lovable Audit Stinger?

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 Lovable Audit Stinger?

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

Is the Lovable Audit Stinger AI skill free?

Yes. It is published on GitHub by legioncodeinc under the AGPL-3.0-or-later 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 👇