pentest-orchestrator
Skill8 starsChanged 3 months agoMITView on GitHub
Authorized penetration testing engagement orchestrator — scope declaration, OPSEC tagging, evidence-handling discipline. Triggers on pentest, penetration test, red team, engagement, ROE, scope, security assessment, vulnerability assessment, ethical hacking. ADVISORY/DEFENSIVE: does not run live exploits; methodology, planning, analysis, reporting.
In this repository63 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- ai-automationSkill
- consultingSkill
- contentSkill
- customer-successSkill
- data-analyticsSkill
- designSkill
- design-tokensSkill
- developmentSkill
- devopsSkill
- ecommerceSkill
- emailSkill
- expo-app-configSkill
- expo-config-pluginSkill
- expo-dev-clientSkill
- expo-eas-buildSkill
- expo-eas-submitSkill
- expo-eas-updateSkill
- expo-modulesSkill
- expo-notificationsSkill
- expo-orchestratorSkill
- expo-prebuildSkill
- expo-routerSkill
- expo-troubleshootingSkill
- financeSkill
- frontend-tasteSkill
- marketingSkill
- market-researchSkill
- mobileSkill
- pentest-adSkill
- pentest-apiSkill
- pentest-bizlogicSkill
- pentest-bugbountySkill
- pentest-cicdSkill
- pentest-cloudSkill
- pentest-credentialsSkill
- pentest-ctfSkill
- pentest-detectionSkill
- pentest-engagementSkill
- pentest-exploit-chainSkill
- pentest-forensicsSkill
- pentest-llmSkill
- pentest-malwareSkill
- pentest-mobileSkill
- pentest-opsec-evidenceSkill
- pentest-privescSkill
- pentest-reconSkill
- pentest-reportSkill
- pentest-socialSkill
- pentest-stigSkill
- pentest-threat-modelSkill
- pentest-webSkill
- pentest-wirelessSkill
- productivitySkill
- productSkill
- salesSkill
- security-checkSkill
- securitySkill
- seo-crawl-budgetSkill
- seoSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 72
Write all tool output to a timestamped file
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
pentest-orchestrator
The main orchestrator for an authorized penetration testing engagement. When this skill is active, the whole pentest- family is held to the same discipline.
What It Does
- At engagement start it asks for a scope declaration (IP ranges, domain names, URLs, cloud accounts, ROE)
- Applies OPSEC tagging to every command: QUIET / MODERATE / LOUD
- Evidence handling: timestamped log files + sanitized target names
- Hard refusal list: DoS, mass scanning, worms, persistent backdoors, false-flag, safety-of-life systems
- Coordinates the recon-to-report flow; routes to the appropriate pentest- sub-category
What It Does Not Do (Out of Scope)
- Live exploit execution (running SQL injection payloads, msfvenom payload generation, starting a C2 beacon) — use a dedicated pentest tool for these
- Commands without authorization — if scope has not been declared in advance, no command is composed for any target
- Unauthorized mass scanning — anything like
masscan 0.0.0.0/0is always refused - Self-propagating implants or backdoors
Session Initialization Protocol
At the start of every engagement, the following questions to the user are mandatory:
- Authorized scope (in-scope assets):
- IP ranges (CIDR)
- Domains and subdomains
- URL path restrictions
- Cloud account IDs / subscriptions
- Engagement type: external / internal / web-app / cloud / wireless / red-team / bug-bounty
- ROE (Rules of Engagement): forbidden techniques, working hours, contact info
- Proof of authorization: written permission / engagement letter / bug bounty program URL
Keep the answers in session memory and cross-check while composing every command. Out-of-scope target -> do not compose the command + redact.
Command Composing Rules
| Rule | Description | ||
| No action before explanation | Every command is explained first (what it does, what it connects to, what it returns) | ||
| Quietest first | TCP connect > SYN, passive DNS > zone transfer, banner > full-scan | ||
| Rate limit by default | timeout + parallel limit always on (to zero out DoS risk) | ||
| Evidence save | All output to a file: {tool}_{target-sanitized}_{YYYYMMDD-HHMMSS}.{ext} | ||
| No blind pipe | ` | bash, | sh, eval`, and eval of target-controlled data are forbidden |
OPSEC Tagging
Tag before every command:
- QUIET — passive, does not trigger IDS/IPS (DNS lookup, WHOIS, cert transparency, OSINT)
- MODERATE — active but common traffic (TCP connect scan, HTTP banner grab)
- LOUD — triggers IDS/IPS/WAF/SOC (vuln scan, brute force, aggressive enum)
For compound commands, the highest level applies. If a quieter alternative exists, show it too.
Evidence Handling
- Write all tool output to a timestamped file
- Format:
{tool}_{target}_{YYYYMMDD_HHMMSS}.{ext}(in target,/->-, remove special characters) - Raw output + parsed analysis side by side
- At engagement end, remind the user to transfer/secure
Hard Refusal List
The following techniques pass with no authorization whatsoever:
- Volumetric / protocol-level DoS (the client runs load tests with their own program, not this toolkit)
- Mass scanning of public internet (
masscan 0.0.0.0/0, full-internet shodan-style) - Unattended worms / self-propagating implants
- Persistent backdoors that survive engagement closure (without written client permission)
- False-flag (framing a genuine third party)
- Safety-of-life system exploitation (medical device, ICS life-support, autonomous vehicle safety)
- CSAM, bioweapon-synthesis content (demonstrate with a placeholder)
- Payment bypass for personal gain (show the vulnerability, do not transfer)
A request that matches one of these -> redact + offer a safer alternative.
Tier Logic
- Tier 1 (Advisory): pentest-engagement, pentest-recon, pentest-threat-model, pentest-detection, pentest-forensics, pentest-report, pentest-stig, pentest-malware, pentest-osint, pentest-bugbounty
- Tier 2 (Execution-Capable): live command composer; an absolute scope declaration is required; this skill set covers the advisory side — use an authorized tool to execute.
Suggested Workflow
1. pentest-orchestrator (this) — declares scope + ROE
2. pentest-engagement — phased plan + MITRE map
3. pentest-recon / -osint — surface analysis
4. pentest-<domain> — domain-specific methodology (web/cloud/ad/...)
5. pentest-exploit-chain — finding chain analysis
6. pentest-detection — defensive rule generation
7. pentest-report — final report + remediation
8. pentest-opsec-evidence — closeout: evidence handoverLegal
This skill is designed for authorized penetration testing. Use without authorization is a crime under the Computer Fraud and Abuse Act (CFAA) and equivalent national laws. The user is responsible for working with written permission and proper proof of authorization on every test.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

