claude-osint
August 30, 2026 · View on GitHub
claude-osint
8 Claude skills · 100+ recon capabilities · 80 secret-regex patterns · 80+ dorks · 9 read-only credential validators · 27 attack-path templates · ~10,000 lines of structured tradecraft. Drop-in
SKILL.mdfiles that turn Claude into a god-mode external recon operator for authorized red-team and bug-bounty engagements.
Built by Sachin Sharma — GenAI Security Research.
SPONSORED BY
What is this?
claude-osint is a library of skills for the Claude skills system. Each skill is a structured SKILL.md file that primes Claude with expert-level methodology for one part of the offensive recon problem.
The core pair — the recon backbone:
osint-methodology- how to think. Strategic + procedural. Asset-graph discipline, severity rubric, time budgeting, identity-fabric mapping, deliverable templates.offensive-osint- what to reach for. Tactical arsenal. Probe paths, regexes, payloads, scoring rules, curl one-liners, tool URLs.
Six organization-grade depth skills — enterprise-scale attack-surface reasoning the core pair doesn't carry:
org-attack-surface- legal entity → owned footprint (GLEIF org-tree, org-first RIR "dark netblock" recall, ASN hyperscaler-scope guard). Discover-only.email-domain-security- composite spoofability verdict (envelope vs header-From; SPF-allalone ≠ spoof-proof) + SPF supply-chain analysis.exposure-risk-quantification- FAIR 0–100 + A–F risk score, $-denominated loss model, board one-pager, ownership/proof honesty caps.continuous-exposure-monitoring- re-scan/diff loop + CTI/ransomware chatter + finding-lifecycle FP discipline + durable alert outbox.cloud-saas-exposure- bucket ownership-gated severity + offline AWS account-ID decode + dependency-confusion confirmation + K8s/CI fingerprint.identity-provider-recon- tenant/federation mapping + pre-auth user-enumeration oracle + name×pattern login synthesis, with a hard enumeration boundary.
Drop them into your Claude environment and it behaves like a senior recon analyst: it knows the techniques, the tooling, the edge cases, and the escalation paths — and it stays in scope.
~10,000 lines of structured tradecraft · 56/56 PASS on a 56-prompt self-evaluation (100%, zero fabrications) · ~85–90% practitioner coverage for the recon phase of authorized engagements.
Structure
claude-osint/
├── skills/
│ ├── osint-methodology/SKILL.md # how to think (515 lines)
│ ├── offensive-osint/
│ │ ├── SKILL.md # what to reach for (4,555 lines)
│ │ ├── scripts/secret_scan.py # stdlib-only 80-pattern secret scanner
│ │ └── scripts/h1_reference.py # HackerOne disclosed-reports reference agent
│ ├── org-attack-surface/SKILL.md # legal entity → owned footprint (1,052 lines)
│ ├── email-domain-security/SKILL.md # spoofability + SPF supply-chain (589 lines)
│ ├── exposure-risk-quantification/SKILL.md # FAIR $ risk score + board one-pager (748 lines)
│ ├── continuous-exposure-monitoring/SKILL.md # re-scan/diff + CTI chatter (811 lines)
│ ├── cloud-saas-exposure/SKILL.md # cloud + supply-chain depth (802 lines)
│ └── identity-provider-recon/SKILL.md # tenant/federation + user-enum (968 lines)
├── docs/ # architecture · coverage · install · usage
├── examples/ # 4 end-to-end engagement walk-throughs
├── tests/smoke-test-prompts.md # 56-prompt self-evaluation
└── assets/banner.png
Each skill directory is self-contained (its own SKILL.md + README.md). Drop into ~/.claude/skills/ and Claude auto-triggers on relevant phrases.
Skill Index
100+ capabilities across 13 domains. Pick a domain to drill in.
Organization-Grade Depth
The six skills that lift the library from single-target recon to enterprise attack-surface reasoning.
| Capability | Skill |
|---|---|
Legal-entity → owned-footprint attribution (GLEIF org-tree by exact LEI · EDGAR / OpenCorporates / Wikidata · reverse-WHOIS · crt.sh ?O= · org-first RIR "dark netblock" recall · ASN hyperscaler-scope guard · discover-only) | org-attack-surface |
Composite email spoofability verdict (envelope vs header-From · SPF -all ≠ spoof-proof · only DMARC governs) + SPF supply-chain (RFC 7208 >10-lookup PermError · dead-include takeover) | email-domain-security |
| FAIR risk quantification (0–100 + A–F score · $-loss IBM/Ponemon per-record bands · board one-pager · ownership/proof demotion cap) | exposure-risk-quantification |
| Continuous exposure monitoring (baseline → re-scan → diff → threshold alert · CTI / ransomware-leak-site chatter · finding-lifecycle FP discipline · durable alert outbox) | continuous-exposure-monitoring |
| Cloud & supply-chain depth (bucket ownership-gated severity · offline AWS account-ID base32 decode · dependency-confusion confirmation · K8s / CI control-plane fingerprint) | cloud-saas-exposure |
| Identity-provider recon (domain→tenant + federation map · pre-auth user-enumeration oracle · name×pattern login synthesis · hard enumeration boundary) | identity-provider-recon |
Reconnaissance & Asset Discovery
| Capability | Skill |
|---|---|
| 6-stage external recon pipeline + time-budget profiles (1h / 4h / 1d / 1w) | methodology |
| Subdomain-source stack (crt.sh + 7-source fallback chain when crt.sh 502s) | arsenal |
| Common-prefix subdomain sweep (100+ ordered prefixes, PowerShell + bash) | arsenal |
| Wayback CDX deep mining + legacy-app pivot (.asp/.php/.jsp/.cfm) | arsenal |
| WHOIS / RDAP / historical-WHOIS + reverse-WHOIS pivots | arsenal |
| Public records (OpenCorporates · SEC EDGAR · GSXT · Rusprofile · Companies House) | arsenal |
| Bulk IP → ASN (Cymru / RIPEstat / bgp.tools) | arsenal |
Identity & SSO Mapping
| Capability | Skill |
|---|---|
| Microsoft Entra (Azure AD) tenant fingerprint + GUID extraction | arsenal |
| M365 deep enum (Teams federation · SharePoint · OneDrive · OAuth · device-code phishing) | arsenal |
| Autodiscover IP correlation (passive M365 confirm even when MX wrapped by Mimecast/Proofpoint) | arsenal |
Okta tenant slug + /api/v1/authn user-enum | arsenal |
| ADFS fingerprint + mex endpoint | arsenal |
| Google Workspace OIDC discovery | arsenal |
| Generic OIDC (Auth0 · Keycloak · Ping · OneLogin · Duo) | arsenal |
| SAML metadata (5 paths) | arsenal |
| AWS account-ID extraction from headers + ARN regex | arsenal |
Web Application Attack Surface
| Capability | Skill |
|---|---|
| Swagger / OpenAPI discovery (28 paths) | arsenal |
| GraphQL discovery + introspection POST body (13 paths) | arsenal |
| GraphQL field-suggestion enum (when introspection disabled) + alias batching + depth bypass | arsenal |
| Always-on HTTP checks (15 paths: .git/.env/actuator/heapdump/etc.) | arsenal |
| Missing security header audit (HSTS/CSP/XFO/etc.) | arsenal |
| Endpoint extraction regex tiers (3 tiers) | arsenal |
| Endpoint interest score (0–100 rubric) | arsenal |
| JS deep analysis · sourcemap leakage · internal-host regex | arsenal |
| Subdomain takeover fingerprints (27 providers) | arsenal |
Cloud & Container
| Capability | Skill |
|---|---|
| Cloud bucket arsenal (S3 / GCS / Azure · 6 prefixes × 15 suffixes × 47 stems) | arsenal |
| Cloud-native fingerprints (Lambda URLs · Cloud Run · Azure Functions · Vercel · Netlify · Workers) | arsenal |
| Kubernetes / etcd / kubelet exposure (12 ports + probes) | arsenal |
| Container registry leak hunting (Docker Hub · Quay · GHCR · ECR · GCR · ACR) | arsenal |
| CI/CD platform exposure (Jenkins · GitLab · TeamCity-KEV · Argo CD · Spinnaker · CircleCI) | arsenal |
Secret & Credential Hunting
| Capability | Skill |
|---|---|
80-pattern secret-regex catalog (48 core + 32 provider-expansion) — §17 ↔ secret_scan.py ↔ §48 kept byte-identical | arsenal |
| Modern AI API keys (Anthropic / OpenAI / HuggingFace / Cloudflare) | arsenal |
| Package-registry & provider tokens (npm / PyPI / Docker Hub / GitLab / Stripe / Shopify / Atlassian / +25 more) | arsenal |
| GitHub code-search dorks (13 templates) | arsenal |
| 9 read-only credential validators (Postman / AWS / GitHub / Slack / Anthropic / OpenAI / npm / Atlassian / DataDog) | arsenal |
| Post-discovery enumeration workflows (IAM enum · repo enum · workspace enum · JWT triage) | arsenal |
secret_scan.py runnable helper (stdlib-only, JSONL output) | arsenal |
h1_reference.py — HackerOne disclosed-reports reference agent (no API key, top-voted / top-bounty / keyword / program filter) | arsenal |
| 80+ dork corpus across 9 categories | arsenal |
Breach Intelligence
| Capability | Skill |
|---|---|
| HudsonRock Cavalier direct API (free; FYI: web-UI wraps a public JSON endpoint) | arsenal |
| Domain-level breach severity mapping | arsenal |
| `SSO_EXPOSURE$ \text{finding} + \text{legacy}-\text{mail}-\text{decommissioned} \text{escalation} \text{pattern} | \text{arsenal} |
| \text{Breach} \times \text{identity} \text{correlation} (\text{HudsonRock} + \text{HIBP} + \text{DeHashed} + \text{IntelX}) | \text{methodology} |
\text{Vendor} & \text{Edge}-\text{Appliance} \text{Fingerprinting}
| \text{Capability} | \text{Skill} |
|---|---|
| \text{Citrix} \text{Netscaler} · \text{F5} \text{BIG}-\text{IP} · \text{Pulse} \text{Secure} / \text{Ivanti} · \text{FortiGate} | \text{arsenal} |
| \text{PaloAlto} \text{GlobalProtect} · \text{Cisco} \text{AnyConnect} · \text{VMware} \text{vCenter} / \text{ESXi} / \text{Horizon} | \text{arsenal} |
| \text{Microsoft} \text{Exchange} \text{OWA} (\text{ProxyShell} / \text{ProxyLogon} / \text{ProxyNotShell}) | \text{arsenal} |
| \text{KEV} \text{CVE} \text{enrichment} + \text{EPSS} \text{scoring} + \text{Metasploit} \text{availability} | \text{arsenal} |
| \text{WAF} / \text{CDN} \text{bypass} + \text{origin} \text{discovery} (8 \text{techniques}) | \text{methodology}, \text{arsenal} |
\text{Email} \text{Security}
| \text{Capability} | \text{Skill} |
|---|---|
| \text{SPF} / \text{DMARC} / \text{DKIM} / \text{BIMI} / \text{MTA}-\text{STS} / \text{TLS}-\text{RPT} / \text{DNSSEC} \text{audit} (\text{bash} + \text{PowerShell}) | \text{arsenal} |
| \text{DMARC} \text{reporting}-\text{vendor} \text{inference} (\text{Kratikal} / \text{dmarcian} / \text{Valimail} / \text{Agari} / \text{EasyDMARC}) | \text{arsenal} |
| \text{TXT} \text{verification} \text{token} \text{catalog} (35+ \text{SaaS} \text{tenants}) | \text{arsenal} |
| \text{MX} → \text{IdP} / \text{mail}-\text{host} \text{inference} | \text{arsenal} |
\text{Human} \text{Intelligence}
| \text{Capability} | \text{Skill} |
|---|---|
| \text{LinkedIn} \text{employee} \text{enumeration} (\text{P0}–\text{P5} \text{role} \text{tiers} · \text{sock}-\text{puppet} \text{hygiene}) | \text{arsenal} |
| \text{Job} \text{posting} \text{tech}-\text{stack} \text{analysis} (\text{Lever} · \text{Greenhouse} · \text{AshbyHQ} · \text{Workable}) | \text{arsenal} |
| \text{Slack} / \text{Discord} / \text{Telegram} / \text{Mattermost} \text{workspace} \text{discovery} | \text{arsenal} |
| \text{Sat} \text{imagery} \text{for} \text{physical} \text{recon} (\text{Google} \text{Earth} · \text{NearMap} · \text{Sentinel} \text{Hub}) | \text{arsenal} |
| \text{Email}-\text{pattern} \text{inference} (8 \text{templates}) | \text{arsenal} |
\text{Supply} \text{Chain}
| \text{Capability} | \text{Skill} |
|---|---|
| \text{Package}-\text{registry} \text{leak} \text{hunting} (\text{npm} · \text{PyPI} · \text{RubyGems} · \text{Cargo} · \text{Packagist} · \text{NuGet} · \text{Maven}) | \text{arsenal} |
| \text{Typosquat} \text{surveillance} | \text{arsenal} |
| \text{Postman} \text{public}-\text{workspace} \text{search} (\text{verified} \text{endpoint}) | \text{arsenal} |
| \text{Stack} \text{Exchange} \text{OSINT} \text{sweep} (8 \text{sites}) | \text{arsenal} |
\text{Reporting} & \text{Deliverables}
| \text{Capability} | \text{Skill} |
|---|---|
| \text{Findings} \text{rubric} (\text{CRITICAL}/\text{HIGH}/\text{MED}/\text{LOW}/\text{INFO} + \text{escalation}) | \text{methodology} |
| \text{Severity} \text{decision} \text{matrix} (88 \text{worked} \text{examples}) | \text{arsenal} |
| \text{Attack}-\text{path} \text{hint} \text{patterns} (27 \text{templates}) | \text{arsenal} |
| \text{Bug}-\text{bounty} \text{submission} \text{templates} (\text{HackerOne} / \text{Bugcrowd} / \text{Intigriti}) | \text{methodology} |
| \text{Client} \text{deliverable} \text{templates} (\text{exec} \text{summary} · \text{risk}-\text{translation} \text{matrix} · \text{cadence}) | \text{methodology} |
| \text{Reproduction} \text{package} | \text{methodology} |
\text{Sector}-\text{Specific}
| \text{Capability} | \text{Skill} |
|---|---|
| \text{Healthcare} (\text{DICOM} · \text{HL7} \text{v2} · \text{FHIR} · \text{Epic} / \text{Cerner} / \text{Allscripts}) | \text{arsenal} |
| \text{Finance} (\text{SWIFT} · \text{FIX} · \text{Bloomberg} · \text{Temenos} / \text{Finacle} / \text{FIS} / \text{Fiserv}) | \text{arsenal} |
| \text{ICS} / \text{SCADA} (\text{Modbus} · \text{BACnet} · \text{Siemens} \text{S7} · \text{DNP3} · \text{EtherNet}/\text{IP}) | \text{arsenal} |
| \text{IoT} (\text{MQTT} · \text{CoAP} · \text{UPnP} · \text{Hikvision} / \text{Dahua} \text{DVRs}) | \text{arsenal} |
\text{Government} ($.gov/.mil` · FedRAMP · FISMA · CUI · SAM.gov) | arsenal |
Capability Map
Eight skills, thirteen capability domains — the core recon pair plus six organization-grade depth skills. Drill into the Skill Index above for concrete sub-capabilities.
%%{init: {'theme':'base', 'themeVariables': {'primaryColor':'#1e293b','primaryTextColor':'#f1f5f9','primaryBorderColor':'#475569','lineColor':'#94a3b8'}}}%%
flowchart LR
Root(["🦅 claude-osint"])
Root --> Core["🎯 Core recon pair"]
Root --> Depth["🏢 Organization-grade depth"]
Core --> M["📘 osint-methodology<br/><i>how to think</i>"]
Core --> A["🛠️ offensive-osint<br/><i>what to reach for</i>"]
Depth --> O["🏢 org-attack-surface<br/><i>entity → footprint</i>"]
Depth --> E["✉️ email-domain-security<br/><i>spoofability + SPF chain</i>"]
Depth --> R["📊 exposure-risk-quantification<br/><i>FAIR $ score</i>"]
Depth --> CM["🔁 continuous-exposure-monitoring<br/><i>watch + diff + chatter</i>"]
Depth --> CL["☁️ cloud-saas-exposure<br/><i>cloud + supply-chain</i>"]
Depth --> ID["🪪 identity-provider-recon<br/><i>tenant + user-enum</i>"]
M --> M1[Recon Pipeline]
M --> M2[Asset Graph]
M --> M3[Identity Fabric]
M --> M4[Findings Rubric]
M --> M5[Reporting Templates]
M --> M6[OpSec & Detectability]
A --> A1[Probe Wordlists]
A --> A2[Vendor Fingerprints]
A --> A3[Cloud · K8s · CI-CD]
A --> A4[Secret Catalog]
A --> A5[Read-Only Validators]
A --> A6[Email Security]
A --> A7[Human Intel]
A --> A8[Sector Notes]
style Root fill:#dc2626,stroke:#7f1d1d,color:#fff
style M fill:#1e293b,stroke:#475569,color:#f1f5f9
style A fill:#7c2d12,stroke:#9a3412,color:#fef3c7
style M1 fill:#0f172a,stroke:#334155,color:#cbd5e1
style M2 fill:#0f172a,stroke:#334155,color:#cbd5e1
style M3 fill:#0f172a,stroke:#334155,color:#cbd5e1
style M4 fill:#0f172a,stroke:#334155,color:#cbd5e1
style M5 fill:#0f172a,stroke:#334155,color:#cbd5e1
style M6 fill:#0f172a,stroke:#334155,color:#cbd5e1
style A1 fill:#1c1917,stroke:#44403c,color:#fed7aa
style A2 fill:#1c1917,stroke:#44403c,color:#fed7aa
style A3 fill:#1c1917,stroke:#44403c,color:#fed7aa
style A4 fill:#1c1917,stroke:#44403c,color:#fed7aa
style A5 fill:#1c1917,stroke:#44403c,color:#fed7aa
style A6 fill:#1c1917,stroke:#44403c,color:#fed7aa
style A7 fill:#1c1917,stroke:#44403c,color:#fed7aa
style A8 fill:#1c1917,stroke:#44403c,color:#fed7aa
style Core fill:#334155,stroke:#475569,color:#f1f5f9
style Depth fill:#134e4a,stroke:#0f766e,color:#ccfbf1
style O fill:#042f2e,stroke:#115e59,color:#99f6e4
style E fill:#042f2e,stroke:#115e59,color:#99f6e4
style R fill:#042f2e,stroke:#115e59,color:#99f6e4
style CM fill:#042f2e,stroke:#115e59,color:#99f6e4
style CL fill:#042f2e,stroke:#115e59,color:#99f6e4
style ID fill:#042f2e,stroke:#115e59,color:#99f6e4
Engagement Flow
%%{init: {'theme':'base', 'themeVariables': {'primaryColor':'#1e293b','primaryTextColor':'#f1f5f9','primaryBorderColor':'#475569','lineColor':'#94a3b8'}}}%%
flowchart TD
A["🎯 Target authorized<br/><i>RoE / BB scope / ASM contract</i>"] --> B[methodology<br/>scope check]
B --> C[methodology<br/>6-stage pipeline]
C --> D1["🔍 Stage 1<br/>Seed Discovery"]
C --> D2["🌐 Stage 2<br/>Asset Expansion"]
C --> D3["📊 Stage 3<br/>Enrichment"]
C --> D4["⚠️ Stage 4<br/>Exposure Analysis"]
C --> D5["📋 Stage 5-6<br/>Converge + Report"]
D1 --> E1[DNS catalog · WHOIS / RDAP<br/>public records<br/><b>org-attack-surface</b><br/>entity → owned footprint]
D2 --> E2[subdomain stack<br/>prefix sweep<br/>Wayback CDX]
D3 --> E3[vendor fingerprint · infra OSINT<br/><b>identity-provider-recon</b> tenant/enum<br/><b>email-domain-security</b> spoofability]
D4 --> E4[secret catalog · HTTP checks<br/>breach × identity · validators<br/><b>cloud-saas-exposure</b> buckets/acct-id/dep-confusion]
D5 --> E5[severity rubric · BB submission<br/><b>exposure-risk-quantification</b> FAIR $ + board<br/><b>continuous-exposure-monitoring</b> re-scan/diff]
E1 --> F[methodology<br/>asset graph]
E2 --> F
E3 --> F
E4 --> G["📋 Findings<br/>severity + confidence + evidence"]
E5 --> H["📦 Deliverable<br/>exec summary · $ risk · repro package"]
F --> G
style A fill:#3b82f6,color:#fff
style B fill:#7c2d12,color:#fef3c7
style C fill:#1e293b,color:#f1f5f9
style F fill:#7c3aed,color:#fff
style G fill:#dc2626,color:#fff
style H fill:#14532d,color:#dcfce7
Usage
With Claude Code
# Install all 8 skills (one-time, after clone)
git clone https://github.com/elementalsouls/Claude-OSINT.git
cd Claude-OSINT
chmod +x ./scripts/sync-skill-content.sh
./scripts/sync-skill-content.sh
mkdir -p ~/.claude/skills
cp -r skills/* ~/.claude/skills/
ls ~/.claude/skills/
Then, in any Claude Code session, ask an OSINT question — the skills auto-load and each triggers on its own relevant phrases. Install only the core pair (osint-methodology + offensive-osint) if you want the recon backbone without the org-grade depth skills.
With the Claude Skills System
# Point Claude at a single skill before starting your session
cat skills/offensive-osint/SKILL.md | claude --system-file -
Manual (Claude.ai / Claude API)
Paste the contents of any SKILL.md into a Project's system prompt or prepend it to your conversation. Both files are plain Markdown — also usable as a personal cheat-sheet without Claude.
Authorization
These skills are intended for assets you own or have written authorization to assess (red-team rules of engagement, bug-bounty in-scope assets, ASM contracts).
All eight skills include a soft scope-check when you ask Claude to act against an unverified third-party target, and the depth skills carry explicit hard-boundary sections (e.g. identity-provider-recon ends at user enumeration — never password spray or credential submission; cloud-saas-exposure decodes an AWS account ID offline but never calls AWS APIs with it). They explicitly exclude active exploitation, post-exploitation, credential submission, token forging/replay, malware development, and other activities beyond OSINT-driven reconnaissance. See SECURITY.md for the full posture.
Documentation
| Doc | Contents |
|---|---|
docs/architecture.md | Design philosophy · asset-graph model · confidence/severity/detectability models · sidecar coordination · diagrams |
docs/coverage.md | Honest practitioner-coverage breakdown by archetype + engagement phase |
docs/installation.md | Symlink installs and multi-environment install patterns |
docs/usage.md | Trigger-phrase reference and prompt templates |
examples/ | 4 end-to-end engagement walk-throughs (quick recon · bug-bounty · M365 deep · secret hunting) |
tests/smoke-test-prompts.md | 56-prompt self-evaluation suite (current grade: 56/56 PASS) |
CHANGELOG.md | Version history |
CONTRIBUTING.md | Pull-request guidelines |
Sponsors
Atlas Cloud is a full-modal AI inference platform that gives developers a single AI API to access video generation, image generation, and LLM APIs. Instead of managing multiple vendor integrations, you connect once and get unified access to 300+ curated models across all modalities.
Check out Atlas Cloud's new coding plan promotion for more budget-friendly API access: https://www.atlascloud.ai/console/coding-plan
ThreatWatch360 is an AI-powered offensive-security platform: continuous Attack Surface Management, AI-assisted penetration testing, brand protection, dark-web monitoring, and Cyber Threat Intelligence in one platform. Validated findings with proof-of-concept and business-impact prioritization — signal over alert fatigue.
About
Operational tradecraft accumulated across external attack-surface engagements, codified into Claude skills. Engagement-platform agnostic - slot into any ASM / ticketing / asset-graph platform you already use, or none.
Author: ElementalSoul
Original framework: SnailSploit/offensive-checklist (v1.x)
Inspired by: Bellingcat's Online Investigations Toolkit · IntelTechniques · OSINT Framework
Tool inventory: . ProjectDiscovery · Six2dez reconftw · SecLists · Assetnote Wordlists
License
This project is dual-licensed:
- Code (Python helpers, shell scripts) — MIT.
- Content (skill methodology, wordlists, regex catalogs, severity rubrics, docs) — Creative Commons Attribution 4.0 International (CC BY 4.0).
To be precise about the boundary: *.py, *.sh, and other source files are MIT; *.md documentation, wordlists, regex catalogs, and rubrics are CC BY 4.0.
Reuse the content freely — including commercially — as long as you credit Sachin Sharma / Claude-OSINT and link back to this repository.
"Give Claude the right skill and it stops being a chatbot. It becomes an operator."