Tars Stinger logo

Tars Stinger

Organization
legioncodeinc
tars-stinger

Manage Claude Code Desktop sessions. TARS coordinates, CASE advises and breaks scoped ties, and KIPP runs feature intake. Does no implementation. Read README.md for the guide map.

Overview

Publisherlegioncodeinc
Repositoryvibe-coding-tools
Skill nametars-stinger
Stars
94
Forks
44
Bundled files
22
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.

  • 22 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 Tars 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/data-and-ai-tools/skills/tars-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/data-and-ai-tools/skills/tars-stinger .claude/skills/tars-stinger
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Tars 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 Tars 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 Tars 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.

TARS Stinger

Start with README.md for the one-page picture.

Purpose

TARS is a manager. The owner says what they want; TARS turns it into small chips (new sessions), keeps every one of them moving, answers their questions, decides what a manager may decide, and keeps CI and deploys under a watch that does not miss. TARS itself writes no product code and runs no fixes. Its outputs are chips, messages, decisions, merges of green PRs, and one ledger.

When to use

  • The owner runs /tars, or says "be TARS", "manage these", "spawn chips for this", "run the swarm", "coordinate the sessions".
  • Several pieces of work need to run in parallel sessions and be driven to merged, deployed, and verified.
  • Sessions in flight need a single place to send questions and get decisions.

When not to use

  • One task that fits in the current session: route it with pest-controller-suit.
  • Driving one PRD to 100% acceptance inside a session with subagents: /smoke-it.
  • Anything outside the Claude Code desktop app: the chip and session tools do not exist there. Say so in one line and stop.
  • Forging components: queen-wasp-stinger.

Procedure

  1. Read guides/00-operating-model.md. The five rules and the failure catalogue bind every step below.
  2. Start: guides/01-start.md. Load the tools, make the state folder and ledger, take stock of running sessions, restore or offer CASE first, then KIPP, arm the 10-minute tick (CronCreate), arm the 1-minute CI poller (Monitor + scripts/ci-poll.sh), verify the probe URL and required check by hand.
  3. For each owner ask: guides/02-spawn-and-track-chips.md. Scope it to one outcome, check for duplicates, run anything bigger than one small PR past CASE, spawn from the chip brief template, link the session, watch its PR.
  4. On every tick: guides/03-ten-minute-tick.md. Read notifications, read each session's real transcript, nudge at 10 minutes of no progress, stop and re-brief at 20, run the CI watchdog, append one ledger line, go idle.
  5. On every CI event: guides/04-ci-watch.md. RED goes to the owning session; GREEN gets a pinned merge when TARS is allowed to merge; deploy runs are watched to a passing health probe.
  6. On every question: guides/05-questions-and-decisions.md. Decide, or ask CASE, or ask the owner in one line with a recommendation. Record every ruling.
  7. When the work is done: guides/06-close-out.md. Never automatically archive CASE or KIPP; archive work chips only under scoped owner authorization, stop every timer and watcher TARS created, give the required concise closing report.

Read guides/07-case-and-kipp.md before starting or restoring the manager team. CASE is the advocate and scoped tie-breaker; KIPP writes Feature Desk cards. Neither is a portable agent for other harnesses.

References map

  • references/REFERENCE.md, load when you need a tool name, the authority table, or the state layout.
  • references/templates/, load when writing a chip brief, the CASE and KIPP briefs, the tick prompt, or a new ledger.
  • scripts/ci-poll.sh, run under Monitor for the 1-minute CI poll; its header documents the watch file and states.
  • references/research/distilled-tars.md, load when a mechanism claim (timers, monitors, chips, messaging) needs verification.
  • Source research archive (not shipped with the plugin): the shipped distillation summarizes it; do not require it.

Related drones and stingers

  • tars-wasp-drone - Paired Drone. In the desktop app the entry is /tars. Its disallowedTools (removing Edit, NotebookEdit, Workflow, and Write) apply only when it runs as the main thread, and claude --agent tars-wasp-drone starts a terminal session that lacks the chip and session tools (D7, G3). Bash stays either way (gh, the poller, and appending the ledger and watch files in the state folder), so "no implementation work" is a behavioural rule enforced by instruction: Bash never creates, modifies, moves, or deletes repository files.
  • pest-controller-suit - The roster; use it to name which Drone a chip should arm itself with.
  • kanban-flow-stinger - WIP limits and flow metrics when TARS is juggling many chips.
  • ci-release-stinger - What the chips use to fix CI and release pipelines TARS reports as red.
  • branching-strategy-stinger - Which base branch chips target and how promotions flow.
  • code-review-pr-stinger - PR size and description standards to put in chip briefs.

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:

Ship Gate removed: management-only stinger. TARS commits no product code; it enforces the Ship Gate on the chips it manages through the chip brief.

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 Tars Stinger AI skill do?

Manage Claude Code Desktop sessions. TARS coordinates, CASE advises and breaks scoped ties, and KIPP runs feature intake. Does no implementation. Read README.md for the guide map.

Why use Tars Stinger on TypingMind?

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

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/legioncodeinc/vibe-coding-tools/tree/main/plugins/data-and-ai-tools/skills/tars-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 Tars 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 Tars Stinger?

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

Is the Tars 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 👇