Careful logo

Careful

Community
fagemx
careful

Safety mode. Warns before destructive commands (rm -rf, DROP TABLE, git push -f, force delete). Does NOT restrict file editing scope — use /guard for that.

Overview

Publisherfagemx
Repositorygstack-game
Skill namecareful
Stars
81
Forks
10
Bundled files
1
LicenseMIT
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.

  • 1 bundled files

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

  • Open source

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

Installation

Install the Careful 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/fagemx/gstack-game/tree/main/skills/careful
  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/fagemx/gstack-game.git /tmp/gstack-game
mkdir -p .claude/skills
cp -r /tmp/gstack-game/skills/careful .claude/skills/careful
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Careful 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 Careful 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 Careful 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.

Preamble (run first)

bash
setopt +o nomatch 2>/dev/null || true  # zsh compat
_GD_VERSION="0.5.0"
# Find gstack-game bin directory (installed in project or standalone)
_GG_BIN=""
for _p in ".claude/skills/gstack-game/bin" ".claude/skills/game-review/../../gstack-game/bin" "$(dirname "$(readlink -f .claude/skills/game-review/SKILL.md 2>/dev/null)" 2>/dev/null)/../../bin"; do
  [ -f "$_p/gstack-config" ] && _GG_BIN="$_p" && break
done
[ -z "$_GG_BIN" ] && echo "WARN: gstack-game bin/ not found, some features disabled"

# Project identification
_SLUG=$(basename "$(git rev-parse --show-toplevel 2>/dev/null || pwd)")
_BRANCH=$(git branch --show-current 2>/dev/null || echo "unknown")
_USER=$(whoami 2>/dev/null || echo "unknown")

# Session tracking
mkdir -p ~/.gstack/sessions
touch ~/.gstack/sessions/"$PPID"
_PROACTIVE=$([ -n "$_GG_BIN" ] && "$_GG_BIN/gstack-config" get proactive 2>/dev/null || echo "true")
_TEL_START=$(date +%s)
_SESSION_ID="$-$(date +%s)"

# Shared artifact storage (cross-skill, cross-session)
mkdir -p ~/.gstack/projects/$_SLUG
_PROJECTS_DIR=~/.gstack/projects/$_SLUG

# Telemetry (sanitize inputs before JSON interpolation)
mkdir -p ~/.gstack/analytics
_SLUG_SAFE=$(printf '%s' "$_SLUG" | tr -d '"\\\n\r\t')
_BRANCH_SAFE=$(printf '%s' "$_BRANCH" | tr -d '"\\\n\r\t')
echo '{"skill":"careful","ts":"'$(date -u +%Y-%m-%dT%H:%M:%SZ)'","repo":"'"$_SLUG_SAFE"'","branch":"'"$_BRANCH_SAFE"'"}'  >> ~/.gstack/analytics/skill-usage.jsonl 2>/dev/null || true

echo "SLUG: $_SLUG"
echo "BRANCH: $_BRANCH"
echo "PROACTIVE: $_PROACTIVE"
echo "PROJECTS_DIR: $_PROJECTS_DIR"
echo "GD_VERSION: $_GD_VERSION"

# Artifact summary
_ARTIFACT_COUNT=$(ls "$_PROJECTS_DIR"/*.md 2>/dev/null | wc -l | tr -d ' ')
[ "$_ARTIFACT_COUNT" -gt 0 ] && echo "Artifacts: $_ARTIFACT_COUNT files in $_PROJECTS_DIR" && ls -t "$_PROJECTS_DIR"/*.md 2>/dev/null | head -5 | while read f; do echo "  $(basename "$f")"; done

Shared artifact directory: $_PROJECTS_DIR (~/.gstack/projects/{slug}/) stores all skill outputs:

  • Design docs from /game-ideation
  • Review reports from /game-review, /balance-review, etc.
  • Player journey maps from /player-experience

All skills read from this directory on startup to find prior work. All skills write their output here for downstream consumption.

If PROACTIVE is "false", do not proactively suggest gstack-game skills.

User Sovereignty

AI models recommend. You decide. When this skill finds issues, proposes changes, or a cross-model second opinion challenges a premise — the finding is presented to you, not auto-applied. Cross-model agreement is a strong signal, not a mandate. Your direction is the default unless you explicitly change it.

Public Output Redaction Lite

Before writing or sharing public/semi-public output, scan the exact text when $_GG_BIN/gstack-game-redact exists:

bash
printf '%s' "$OUTPUT_TEXT" | "$_GG_BIN/gstack-game-redact" --json

Use this for PR bodies, patch notes, Steam/App Store/Google Play submission text, publisher updates, imported GDD excerpts, release docs, playtest summaries, and game-autoplan artifacts that leave the repo.

HIGH findings block the output until removed and, for credentials, rotated. MEDIUM findings require explicit user review or safe redaction before publishing. Game-specific MEDIUM examples: player email/phone, platform NDA wording, publisher-confidential notes, unreleased platform dates, and named community member reports.

Completion Status Protocol

DONE / DONE_WITH_CONCERNS / BLOCKED / NEEDS_CONTEXT. Escalation after 3 failed attempts.

Telemetry (run last)

bash
_TEL_END=$(date +%s)
_TEL_DUR=$(( _TEL_END - _TEL_START ))
[ -n "$_GG_BIN" ] && "$_GG_BIN/gstack-telemetry-log" \
  --skill "careful" --duration "$_TEL_DUR" --outcome "OUTCOME" \
  --used-browse "false" --session-id "$_SESSION_ID" 2>/dev/null &

/careful: Destructive Command Safety

Activates heightened awareness for destructive operations. When active, flag and confirm before executing any potentially destructive command.

What triggers a warning

PatternRiskAction
rm -rf (except node_modules, .next, dist, build, pycache)File deletionConfirm before executing
git push -f / git push --forceHistory rewriteConfirm + warn about remote impact
git reset --hardUncommitted work lossConfirm + suggest stash first
git clean -fUntracked file deletionConfirm + list what will be deleted
git branch -DBranch deletionConfirm + check if merged
DROP TABLE / DROP DATABASEData destructionConfirm + verify environment
TRUNCATEData deletionConfirm
docker system pruneContainer cleanupConfirm
Kill/stop commands on game serversService disruptionConfirm + check player count

Safe exceptions (no warning needed)

rm -rf node_modules/
rm -rf .next/
rm -rf dist/
rm -rf build/
rm -rf __pycache__/
rm -rf .cache/
rm -rf tmp/
git push (without -f)

Activation

This skill is activated by invoking /careful. It stays active for the remainder of the session.

When active, before any Bash command that matches a warning pattern:

⚠️ CAREFUL MODE: This command will [describe impact]. Affected: [list files/data/branches] Reversible: [yes/no/partially]

Proceed? (confirm to execute)

Deactivation

Say "turn off careful mode" or start a new session.

Review Log

bash
[ -n "$_GG_BIN" ] && "$_GG_BIN/gstack-review-log" '{"skill":"careful","timestamp":"TIMESTAMP","status":"ACTIVATED","commit":"COMMIT"}' 2>/dev/null || true

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 Careful AI skill do?

Safety mode. Warns before destructive commands (rm -rf, DROP TABLE, git push -f, force delete). Does NOT restrict file editing scope — use /guard for that.

Why use Careful on TypingMind?

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

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/fagemx/gstack-game/tree/main/skills/careful. 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 Careful?

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 Careful?

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

Is the Careful AI skill free?

Yes. It is published on GitHub by fagemx under the MIT 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 👇