injection — Claude Code skill: SQL, XXE, SSTI vulnerability testing
A skill for Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, and OpenCode that tests SQL, NoSQL, OS Command, SSTI, XXE, and LDAP/XPath injection vulnerabilities. Runs from input discovery through exploitation to PoC capture.
MIT 210 installs ★ 632 updated 2026-07-29
npx skills add transilienceai/communitytools --skill injection -a claude-codeWhat injection does
The skill follows a five-step workflow: identify injection points across parameters, headers, cookies, and JSON fields; probe minimally to detect the injection type; apply context-appropriate payloads; escalate toward data extraction or RCE; then capture evidence and write a PoC.
Each attack class has its own technique set. SQL injection covers in-band (union, error-based), blind (boolean, time-based), and out-of-band methods. NoSQL covers operator injection, JavaScript injection, and aggregation pipeline abuse. SSTI includes template engine fingerprinting, sandbox escape, and RCE chains. XXE covers entity expansion, SSRF via XXE, blind XXE, and parameter entities. LDAP and XPath cover filter manipulation and authentication bypass. Reference files install alongside the skill under the reference/ directory.
Because the escalation chain leads to RCE and file reads, this skill is for authorised penetration testing only — running it against systems you do not have permission to test is out of scope.
What it helps with
- test a login form or API endpoint for SQL injection and produce a working PoC
- fingerprint an injection type with the fewest possible probes
- detect SSTI in a template field and attempt sandbox escape
- find blind XXE and check whether SSRF is reachable via entity expansion
- bypass LDAP authentication through filter manipulation
Example prompts for the agent
- Test
/api/users?id=1for SQL injection, covering blind and out-of-band techniques - Check the
templatefield in this POST request for SSTI and try to reach RCE - Run an XXE test on this XML endpoint and determine if SSRF is possible
How to install injection
1Skills CLI (Vercel)
npx skills add transilienceai/communitytools --skill injection -a claude-code2Manually: put the files in place
~/.claude/skills/injection/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/injection/
1Skills CLI (Vercel)
npx skills add transilienceai/communitytools --skill injection -a codex2Manually: put the files in place
.agents/skills/injection/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add transilienceai/communitytools --skill injection -a cursor2Manually: put the files in place
.cursor/skills/injection/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install transilienceai/communitytools skills/injection2Manually: put the files in place
.github/skills/injection/SKILL.md1Skills CLI (Vercel)
npx skills add transilienceai/communitytools --skill injection -a gemini-cli2Platform CLI
gemini skills install https://github.com/transilienceai/communitytools --path skills/injectionManually: put the files in place
.devin/skills/injection/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add transilienceai/communitytools --skill injection -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about injection
How do I install the injection skill in Claude Code?
Run npx skills add transilienceai/communitytools --skill injection -a claude-code. The skill is placed at ~/.claude/skills/injection/SKILL.md.
What does the injection skill do?
It guides the agent through a full injection-testing cycle: finding input vectors, detecting the vulnerability type, exploiting with targeted payloads, and writing a PoC with evidence. It covers SQL, NoSQL, OS Command, SSTI, XXE, and LDAP/XPath.
How do I install the injection skill in Cursor?
Run npx skills add transilienceai/communitytools --skill injection -a cursor. The skill lands at .cursor/skills/injection/SKILL.md.
Source text
SKILL.md · MIT
--- name: injection description: Injection vulnerability testing - SQL, NoSQL, OS Command, SSTI, XXE, and LDAP/XPath injection techniques. --- # Injection Test for injection vulnerabilities across all input vectors. Covers SQL, NoSQL, Command, SSTI, XXE, and LDAP injection. ## Techniques | Type | Key Vectors | |------|-------------| | **SQL Injection** | In-band (union, error), Blind (boolean, time), Out-of-band | | **NoSQL Injection** | Operator injection, JavaScript injection, aggregation pipeline | | **Command Injection** | OS command separators, blind techniques, out-of-band | | **SSTI** | Template engine detection, sandbox escape, RCE chains | | **XXE** | Entity expansion, SSRF via XXE, blind XXE, parameter entities | | **LDAP/XPath** | Filter manipulation, authentication bypass | ## Workflow 1. Identify injection points (parameters, headers, cookies, JSON fields) 2. Detect injection type with minimal probes 3. Exploit with context-appropriate payloads 4. Escalate (data extraction, RCE, file read) 5. Capture evidence and write PoC ## Reference - `reference/sql-injection*.md` - SQL injection techniques - `reference/nosql-injection*.md` - NoSQL injection techniques - `reference/os-command-injection*.md` - OS command injection - `reference/ssti*.md` - Server-side template injection - `reference/xxe*.md` - XML external entity injection - `reference/ldap-injection-quickstart.md` - LDAP filter injection: detection, auth bypass, blind boolean extraction via `(description=PREFIX*)` chaining - `reference/xpath-injection-quickstart.md` - XPath injection (CWE-643): lxml/Java/Node sinks, `' or '1'='1' or 'a'='b` boolean oracle, blind char-by-char extraction recipe
© transilienceai · MIT · transilienceai/communitytools