Closes the 'rules in the box' gap — the README has claimed YARA + Falco coverage but detect_yara and detect_falco were NULL on every module. This commit lights up both formats for the 6 highest-value modules (covering 10 of 31 registered modules via family-shared rules), and the existing operational playbook gains the format-specific deployment recipes + the cross-format correlation table. YARA rules (8 rules, 9 module-headers, 152 lines): - copy_fail_family — etc_passwd_uid_flip + etc_passwd_root_no_password (shared across copy_fail / copy_fail_gcm / dirty_frag_esp / dirty_frag_esp6 / dirty_frag_rxrpc) - dirty_pipe — passwd UID flip pattern, dirty-pipe-specific tag - dirtydecrypt — 28-byte ELF prefix match on tiny_elf[] + setuid+execve shellcode tail, detects the page-cache overlay landing - fragnesia — 28-byte ELF prefix on shell_elf[] + setuid+setgid+seteuid cascade, detects the 192-byte page-cache overlay - pwnkit — gconv-modules cache file format (small text file with module UTF-8// X// /tmp/...) - pack2theroot — malicious .deb (ar archive + SUID-bash postinst) + /tmp/.suid_bash artifact scan Falco rules (13 rules, 9 module-headers, 219 lines): - pwnkit — pkexec with empty argv + GCONV_PATH/CHARSET env from non-root - copy_fail_family — AF_ALG socket from non-root + NETLINK_XFRM from unprivileged userns + /etc/passwd modified by non-root - dirty_pipe — splice() of setuid/credential file by non-root - dirtydecrypt — AF_RXRPC socket + add_key(rxrpc) by non-root - fragnesia — TCP_ULP=espintcp from non-root + splice of setuid binary - pack2theroot — SUID bit set on /tmp/.suid_bash + dpkg invoked by packagekitd with /tmp/.pk-*.deb + 2x InstallFiles on same transaction Wiring: each module's .detect_yara and .detect_falco struct fields now point at the embedded string. The dispatcher dedups by pointer, so family-shared rules emit once across the 5 sub-modules. docs/DETECTION_PLAYBOOK.md augmented (302 -> 456 lines): - New 'YARA artifact scanning' subsection under SIEM integration with scheduled-scan cron pattern + per-rule trigger table - New 'Falco runtime detection' subsection with deploy + per-rule trigger table - New 'Per-module detection coverage' table — 4-format matrix - New 'Correlation across formats' section — multi-format incident signature per exploit (the 3-of-4 signal pattern) - New 'Worked example: catching DirtyDecrypt end-to-end' walkthrough from Falco page through yara confirmation, recovery, hunt + patch The existing operational lifecycle / SIEM patterns / FP tuning content is preserved unchanged — this commit only adds. Final stats: - auditd: 109 rule statements across 27 modules - sigma: 16 sigma rules across 19 modules - yara: 8 yara rules across 9 module headers (5 family + 4 distinct) - falco: 13 falco rules across 9 module headers The remaining 21 modules can gain YARA / Falco coverage incrementally by populating their detect_yara / detect_falco struct fields.
17 KiB
SKELETONKEY detection playbook
Operational guide for blue teams using SKELETONKEY defensively. Pairs
with docs/DEFENDERS.md (the "what" reference) — this is the "how to
make it part of your daily ops" guide.
The lifecycle
┌─────────────┐
│ inventory │ ← skeletonkey --list (what's bundled?)
└──────┬──────┘
▼
┌─────────────┐
│ scan │ ← skeletonkey --scan --json (what am I vulnerable to?)
└──────┬──────┘
▼
┌─────────────┐
│ fleet scan │ ← skeletonkey-fleet-scan.sh hosts.txt
└──────┬──────┘
▼
┌────────────┼────────────┐
▼ ▼ ▼
┌────────┐ ┌─────────┐ ┌──────────┐
│ deploy │ │ mitigate│ │ upgrade │ ← three responses
│ rules │ │ (pre-fix│ │ (kernel │
│(SIEM) │ │ stopgap)│ │ patch) │
└────┬───┘ └─────┬───┘ └─────┬────┘
└────────────┼────────────┘
▼
┌─────────────┐
│ monitor │ ← ausearch -k skeletonkey-* / SIEM alerts
└─────────────┘
Recipes by team size
Single host (workstation / single server)
# Daily/weekly hygiene check
sudo skeletonkey --scan
# If anything's VULNERABLE, deploy detections + apply mitigation
sudo skeletonkey --detect-rules --format=auditd | sudo tee /etc/audit/rules.d/99-skeletonkey.rules
sudo augenrules --load
sudo skeletonkey --mitigate copy_fail # or whichever module fired
Small fleet (~10-100 hosts, SSH-reachable)
Use tools/skeletonkey-fleet-scan.sh:
# Hosts list — one per line; user@host:port supported
cat > hosts.txt <<EOF
prod-web-01
prod-web-02
deploy@bastion-01
ops@db-01:2222
EOF
# Scan; binary scp'd, run, cleaned up. Output is one JSON doc.
./skeletonkey-fleet-scan.sh \
--binary ./skeletonkey \
--ssh-key ~/.ssh/ops_key \
--parallel 8 \
hosts.txt > fleet-scan-$(date +%F).json
# Show me hosts with any VULNERABLE finding
jq '.hosts[] | select(.scan.modules | map(.result == "VULNERABLE") | any) | .host' \
fleet-scan-*.json
# Show summary across the fleet
jq '.summary' fleet-scan-*.json
Output shape:
{
"generated_at": "2026-05-16T22:00:00Z",
"n_hosts": 4,
"summary": {
"ok": 4,
"failed": 0,
"vulnerable": [
{ "cve": "CVE-2024-1086", "name": "nf_tables", "count": 2 },
{ "cve": "CVE-2023-0458", "name": "entrybleed", "count": 4 }
]
},
"hosts": [...]
}
Larger fleet (>100 hosts)
skeletonkey-fleet-scan.sh is intentionally simple (parallel ssh). For
fleets too large for SSH-fan-out, wrap it in your config-management
tool of choice:
- Ansible: ship the binary via
copy:, run viacommand:, parse JSON withjqin a follow-on task - SaltStack:
cmd.runreturning JSON;salt-call --returnto your SIEM - Fabric / Mitogen: same shape, just Python-side
Sample Ansible task:
- name: scan with skeletonkey
copy:
src: skeletonkey
dest: /tmp/skeletonkey
mode: '0755'
- name: run --scan --json
command: /tmp/skeletonkey --scan --json --no-color
register: scan
changed_when: false
failed_when: false # skeletonkey exit codes are semantic, not errors
- name: collect
set_fact:
skeletonkey_scan: "{{ scan.stdout | from_json }}"
- name: cleanup
file:
path: /tmp/skeletonkey
state: absent
SIEM integration patterns
Splunk
# splunk input config (inputs.conf)
[script:///opt/skeletonkey/skeletonkey-cron-scan.sh]
interval = 86400
source = skeletonkey
sourcetype = skeletonkey:scan
skeletonkey-cron-scan.sh:
#!/bin/bash
/usr/local/bin/skeletonkey --scan --json --no-color
Search the indexed events:
index=skeletonkey sourcetype="skeletonkey:scan" modules{}.result=VULNERABLE
| stats count by host modules{}.cve
Elastic / OpenSearch
Filebeat module reading the per-host scan JSON files (one per day),
indexed into an skeletonkey-* index pattern. Standard Kibana
visualization on modules.cve over time tracks vulnerability lifecycle.
Sigma → your platform
# Ship Sigma rules into your platform
skeletonkey --detect-rules --format=sigma > /etc/sigma/skeletonkey.yml
# Convert to your target (Sentinel, Elastic, etc.) via sigmac
sigmac -t elastic /etc/sigma/skeletonkey.yml
YARA artifact scanning
YARA rules catch the post-fire state — page-cache shellcode
overwrites, malicious .deb drops, /etc/passwd UID flips. Run them
as a scheduled scan against sensitive paths:
# Ship YARA rules
sudo skeletonkey --detect-rules --format=yara | sudo tee /etc/yara/skeletonkey.yar
# Scheduled scan via cron — catches the page-cache and /tmp artifacts
# /etc/cron.d/skeletonkey-yara
*/15 * * * * root yara -r /etc/yara/skeletonkey.yar \
/etc/passwd /tmp /usr/bin/su /usr/bin/passwd \
2>>/var/log/skeletonkey-yara.log
What each rule catches:
| Rule | Triggers on |
|---|---|
etc_passwd_uid_flip |
Non-root user line in /etc/passwd with a zero-padded UID (0000+). Canonical Copy Fail / Dirty Frag / Dirty Pipe / DirtyDecrypt outcome. |
etc_passwd_root_no_password |
root line with empty password field — DirtyDecrypt's intermediate corruption step. |
pwnkit_gconv_modules_cache |
Small gconv-modules text file with a module UTF-8// X// /tmp/… redefinition. |
dirty_pipe_passwd_uid_flip |
Same UID-flip pattern (Dirty Pipe-specific tag). |
dirtydecrypt_payload_overlay |
First 28 bytes of /usr/bin/su (or similar) match the embedded 120-byte ET_DYN shellcode the V12 PoC overlays. |
fragnesia_payload_overlay |
Same shape for the 192-byte Fragnesia payload. |
pack2theroot_malicious_deb |
.deb ar-archive in /tmp with the SUID-bash postinst. |
pack2theroot_suid_bash_drop |
/tmp/.suid_bash exists and is a real bash ELF. |
The page-cache overlay rules (dirtydecrypt_payload_overlay,
fragnesia_payload_overlay) are particularly high-signal: no
legitimate ELF starts with those exact 28 bytes, so a hit means the
exploit landed.
Falco runtime detection
Falco catches the exploit as it fires by hooking syscalls and namespace events. Best deploy for K8s / container hosts but works on any modern Linux:
sudo skeletonkey --detect-rules --format=falco \
| sudo tee /etc/falco/rules.d/skeletonkey.yaml
sudo falco --validate /etc/falco/rules.d/skeletonkey.yaml
sudo systemctl reload falco # or restart, depending on distro
What each rule catches:
| Rule | Triggers on |
|---|---|
Pwnkit-style pkexec invocation |
pkexec spawned with empty argv (the bug's hallmark). |
Pwnkit-style GCONV_PATH injection |
Non-root sets GCONV_PATH= / CHARSET= before spawning a setuid binary. |
AF_ALG authenc keyblob installed by non-root |
socket(AF_ALG) by non-root — Copy Fail / GCM variant primitive. |
XFRM NETLINK_XFRM bind from unprivileged userns |
XFRM SA setup from non-root userns — Dirty Frag / Fragnesia primitive. |
/etc/passwd modified by non-root |
Post-fire signal for the whole page-cache-write family. |
Dirty Pipe splice from setuid/sensitive file by non-root |
splice() of /etc/passwd or /usr/bin/su by non-root. |
AF_RXRPC socket created by non-root |
DirtyDecrypt primitive — socket(AF_RXRPC) is nearly unheard-of in production. |
rxrpc security key added |
add_key("rxrpc", …) by non-root — DirtyDecrypt handshake setup. |
TCP_ULP=espintcp set by non-root |
Fragnesia trigger — flipping a TCP socket to espintcp ULP. |
SUID bash dropped to /tmp |
Pack2TheRoot postinst landing /tmp/.suid_bash. |
dpkg invoked by PackageKit on behalf of non-root caller |
Pack2TheRoot chain — packagekitd → dpkg installing a /tmp .pk-*.deb. |
Day-to-day operational shape
What "good" looks like in the SIEM
- Daily
skeletonkey --scan --jsonfrom every host indexed - Trend dashboard: count of VULNERABLE results by CVE over time
- Goal: every VULNERABLE → OK transition within SLA (e.g., 14 days for patched-mainline bugs, 24h for actively-exploited)
- Alert on: any host with a result not seen yesterday (could indicate a config drift, a new install, or a disabled mitigation)
Auditd events from the embedded rules
After deploying skeletonkey --detect-rules --format=auditd:
# By module key
sudo ausearch -k skeletonkey-copy-fail -ts today
sudo ausearch -k skeletonkey-dirty-pipe -ts today
sudo ausearch -k skeletonkey-pwnkit -ts today
sudo ausearch -k skeletonkey-nf-tables-userns -ts today
sudo ausearch -k skeletonkey-overlayfs -ts today
# Anything skeletonkey-tagged in the last hour
sudo ausearch -k 'skeletonkey-*' -ts recent
# Forward to syslog (rsyslog example)
# /etc/rsyslog.d/skeletonkey.conf:
:msg, contains, "skeletonkey-" @@your-siem.example.com:514
When a VULNERABLE result fires
Decision tree:
A scan reports VULNERABLE for module X
│
├── Q: Can I patch the underlying kernel / package?
│ ├── YES → schedule patch window. In the meantime:
│ │ skeletonkey --mitigate X (if supported)
│ │ Verify auditd rule for X is loaded.
│ │ Monitor for the rule key.
│ └── NO (legacy LTS, embedded device, prod freeze) →
│ skeletonkey --mitigate X (essential)
│ Compensating control: tighten LSM (SELinux/AppArmor)
│ Document in risk register
│
└── Q: Was this VULNERABLE before? When?
├── First time → config drift; investigate why detection now
│ produces this result
└── Persistent → mitigation isn't applied OR is being reverted
by config management; fix the config baseline
Mitigation reverts
Mitigations can break legitimate functionality:
| Mitigation | Side effect |
|---|---|
copy_fail blacklist algif_aead |
strongSwan / IPsec breaks |
copy_fail blacklist esp4/esp6 |
IPsec breaks |
copy_fail blacklist rxrpc |
AFS / kAFS clients break |
copy_fail AppArmor restrict userns=1 |
bubblewrap, podman rootless break |
If you applied a mitigation and now need to revert (e.g., the kernel patch has rolled out fleet-wide):
sudo skeletonkey --cleanup copy_fail
# OR manually:
sudo rm /etc/modprobe.d/dirtyfail-mitigations.conf
sudo rm /etc/sysctl.d/99-dirtyfail-mitigations.conf
# Reload affected modules / sysctls per your distro
Per-module detection coverage
Across the 4 rule formats:
| Module | CVE | auditd | sigma | yara | falco |
|---|---|---|---|---|---|
| copy_fail | CVE-2026-31431 | ✓ | ✓ | ✓ | ✓ |
| copy_fail_gcm | (variant) | ✓ | ✓ | ✓ | ✓ |
| dirty_frag_esp | CVE-2026-43284 | ✓ | ✓ | ✓ | ✓ |
| dirty_frag_esp6 | CVE-2026-43284 | ✓ | ✓ | ✓ | ✓ |
| dirty_frag_rxrpc | CVE-2026-43500 | ✓ | ✓ | ✓ | ✓ |
| dirty_pipe | CVE-2022-0847 | ✓ | ✓ | ✓ | ✓ |
| dirtydecrypt | CVE-2026-31635 | ✓ | ✓ | ✓ | ✓ |
| fragnesia | CVE-2026-46300 | ✓ | ✓ | ✓ | ✓ |
| pwnkit | CVE-2021-4034 | ✓ | ✓ | ✓ | ✓ |
| pack2theroot | CVE-2026-41651 | ✓ | ✓ | ✓ | ✓ |
| Other 21 modules | various | ✓ | partial | — | — |
Full 4-format coverage on the 10 highest-value modules; auditd
covers everything. YARA / Falco expansion to the remaining 21 modules
is incremental contributor work (each module's detect_yara /
detect_falco field in the module struct just needs a string).
Correlation across formats
Single-format detections are useful; the high-confidence signal is the correlation across formats for the same module in a short window. Each exploit leaves a recognisable multi-format trail:
| Exploit | falco fires | auditd fires | yara confirms |
|---|---|---|---|
| Pwnkit | pkexec empty argv |
execve /usr/bin/pkexec + GCONV_PATH= env |
gconv-modules cache in /tmp |
| Dirty Pipe | splice() from /etc/passwd |
splice + write to /etc/passwd |
UID flip in /etc/passwd |
| Copy Fail | socket(AF_ALG) |
algif_aead + ALG_SET_KEY |
UID flip in /etc/passwd |
| Dirty Frag (ESP) | NETLINK_XFRM sendto + TCP_ULP | XFRM_MSG_NEWSA | UID flip in /etc/passwd |
| DirtyDecrypt | socket(AF_RXRPC) + add_key(rxrpc) |
AF_RXRPC + add_key | 120-byte ELF overwrites /usr/bin/su |
| Fragnesia | TCP_ULP=espintcp from non-root |
XFRM + setsockopt(TCP_ULP) | 192-byte ELF overwrites /usr/bin/su |
| Pack2TheRoot | dpkg invoked by packagekitd with /tmp/.pk-*.deb | new .deb in /tmp + chmod 4755 on /tmp/.suid_bash |
malicious .deb + SUID bash both present |
If three of the four signals fire for the same module in the same window, the exploit landed. One signal alone in a noisy environment is more likely a tuning FP; three signals is incident response.
Worked example: catching DirtyDecrypt end-to-end
A SOC operator gets a Falco page:
CRITICAL AF_RXRPC socket() by non-root (user=alice proc=poc pid=44231)
-
Confirm via auditd — pull events keyed on the family:
sudo ausearch -k skeletonkey-dirtydecrypt-rxrpc -ts recentExpect:
socket(...,33,...)+ subsequentadd_key("rxrpc",...). -
Confirm via yara — scan setuid binaries for the page-cache overlay:
yara /etc/yara/skeletonkey.yar /usr/bin/su /usr/bin/passwdIf
dirtydecrypt_payload_overlaymatches/usr/bin/su, the exploit landed — the binary's page cache has been overwritten with the 120-byte shellcode. -
Recover — the on-disk binary is intact; only the page cache is corrupted. Drop it:
sudo skeletonkey --cleanup dirtydecrypt # or: echo 3 > /proc/sys/vm/drop_caches -
Sigma hunt for lateral / repeat — query your SIEM with the sigma rule ID
7c1e9a40-skeletonkey-dirtydecryptover the last 7 days to find any other hosts. -
Patch. DirtyDecrypt's mainline fix is commit
a2567217in Linux 7.0 — seeCVES.mdfor distro backports. -
Harden.
rxrpcis rarely needed on non-AFS hosts:echo "blacklist rxrpc" | sudo tee /etc/modprobe.d/blacklist-rxrpc.conf sudo update-initramfs -u
The same shape applies to every module: pick the auditd key, the yara rule for the artifact, the falco rule for the runtime signal, and the sigma rule for the hunt.
Common false positives + tuning
| Rule key | False positive | Fix |
|---|---|---|
skeletonkey-copy-fail-afalg |
strongSwan, libcrypto using kernel crypto | -F auid= exclude service account UIDs |
skeletonkey-dirty-pipe-splice |
nginx, HAProxy, kTLS | -F gid!=33 -F gid!=99 exclude web service accounts |
skeletonkey-pwnkit-execve |
gnome-software, polkit's own re-exec | Correlate by parent process; pkexec via gnome dbus is benign |
skeletonkey-nf-tables-userns |
docker rootless, podman, snap confined apps | Whitelist known userns-using service GIDs |
skeletonkey-overlayfs |
docker / containerd mounting overlayfs as root | The rule is intended for unprivileged-userns overlayfs mounts; add -F auid>=1000 |
Pre-patch quarantine pattern
If a CVE is in active exploitation and you can't patch immediately:
# Stage 1: detect
sudo skeletonkey --scan --json | jq '.modules[] | select(.cve == "CVE-XXXX")'
# Stage 2: mitigate (where supported)
sudo skeletonkey --mitigate <module>
# Stage 3: monitor — auditd rules already deployed
sudo ausearch -k 'skeletonkey-*' -ts today | grep <module>
# Stage 4: contain — temporarily restrict the trigger surface
# e.g., for nf_tables CVE-2024-1086:
echo 0 | sudo tee /proc/sys/kernel/unprivileged_userns_clone
# OR
sudo sysctl -w kernel.apparmor_restrict_unprivileged_userns=1
# Stage 5: alert
# When auditd or sigma rule fires, page on-call
Maintenance contract
When SKELETONKEY ships a new module:
- CI test passes on at least one vulnerable + patched kernel pair
- Detection rules ship alongside (auditd + sigma minimum)
- CVES.md row added with patch status
- NOTICE.md credits original researcher
- ROADMAP.md updated
Treat these as the SLA for any blue-team-facing deliverable.
When you find a new false positive
File an issue at https://github.com/KaraZajac/SKELETONKEY/issues with:
- The exact ausearch line that fired
- The legitimate process that produced it
- Distro / kernel version
Most false-positive fixes are a -F filter on the embedded rule —
small, mergeable.