Improve detection for Python setuptools backdoors (chainguard-dev#164)

* Improve reliability of py_setuptools rules * Fix GoReleaser (chainguard-dev#162) * Fix GoReleaser * Add project_name * Remove problematic MS repository from Checks * Repository has been fixed * Import rule URLs, add them to markdown & JSON output (chainguard-dev#165) * Add URLs to markdown output * run prettier against output * Disable or modify GitHub actions to work better with third_party code (chainguard-dev#152) * Bump actions/checkout from 4.1.3 to 4.1.4 in the all group (chainguard-dev#171) Bumps the all group with 1 update: [actions/checkout](https://github.com/actions/checkout). Updates `actions/checkout` from 4.1.3 to 4.1.4 - [Release notes](https://github.com/actions/checkout/releases) - [Changelog](https://github.com/actions/checkout/blob/main/CHANGELOG.md) - [Commits](actions/checkout@1d96c77...0ad4b8f) --- updated-dependencies: - dependency-name: actions/checkout dependency-type: direct:production update-type: version-update:semver-patch dependency-group: all ... Signed-off-by: dependabot[bot] <[email protected]> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Add ThreatHunting-Keywords-yara-rules (chainguard-dev#160) * Add ThreatHunting-Keywords-yara-rules * Update README * Remove superfluous mkdir command * Fix tests * Fix tests * Remove merge artifacts * Pin rules to a known commit; add check for updates * Re-add -s * Avoid using jq for portability * Add new commit to output * Fix test * Bump golangci/golangci-lint-action from 4.0.0 to 5.0.0 (chainguard-dev#172) Bumps [golangci/golangci-lint-action](https://github.com/golangci/golangci-lint-action) from 4.0.0 to 5.0.0. - [Release notes](https://github.com/golangci/golangci-lint-action/releases) - [Commits](golangci/golangci-lint-action@3cfe3a4...82d40c2) --- updated-dependencies: - dependency-name: golangci/golangci-lint-action dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <[email protected]> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Thomas Strömberg <[email protected]> Co-authored-by: Ville Aikas <[email protected]> * Improve how we approach Python code * Add 2021.DiscordSafety sample * Make pythonSetup private again * Add aiohttp sample --------- Signed-off-by: dependabot[bot] <[email protected]> Co-authored-by: Evan Gibler <[email protected]> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Ville Aikas <[email protected]>
egibs · May 13, 2024 · 5a95dc0 · 5a95dc0
1 parent bdcd078
commit 5a95dc0
Show file tree

Hide file tree

Showing 14 changed files with 398 additions and 107 deletions.
diff --git a/pkg/report/report.go b/pkg/report/report.go
@@ -400,8 +400,8 @@ func Generate(ctx context.Context, path string, mrs yara.MatchRules, ignoreTags
 			continue
 		}
 
-		// If the existing description is longer,
-		if len(existing.Description) < len(b.Description) {
+		// If the existing description is longer and the priority is the same or lower
+		if len(existing.Description) < len(b.Description) && existing.RiskScore <= b.RiskScore {
 			fr.Behaviors[key].Description = b.Description
 		}
 

diff --git a/rules/combo/backdoor/py_setuptools.yara b/rules/combo/backdoor/py_setuptools.yara
@@ -1,78 +1,69 @@
 import "math"
 
-rule setuptools_cmd_exec : high {
+private rule pythonSetup {
+  strings:
+    $i_distutils = "from distutils.core import setup"
+    $i_setuptools = "setuptools"
+    $setup = "setup("
+  condition:
+    filesize < 2097152 and $setup and any of ($i*)
+}
+
+rule setuptools_cmd_exec : suspicious {
   meta:
     description = "Python library installer that executes external commands"
-    hash_2022_laysound_4_5_2_setup = "4465bbf91efedb996c80c773494295ae3bff27c0fff139c6aefdb9efbdf7d078"
-    hash_2022_2022_requests_3_0_0_setup = "15507092967fbd28ccb833d98c2ee49da09e7c79fd41759cd6f783672fe1c5cc"
-    hash_2023_grandmask_3_13_setup = "8835778f9e75e6493693fc6163477ec94aba723c091393a30d7e7b9eed4f5a54"
   strings:
-    $setup = "setup(" fullword
-    $setuptools = "setuptools"
-    $distutils = "distutils"
-    $s_sys_val = /os.system\([\"\'\w\ \-\)\/]{0,64}/
-    $s_subprocess_val = /subprocess.\w{0,32}\([\"\'\/\w\ \-\)]{0,64}/
-    $s_import = "import subprocess"
+    $os_system = /os.system\([\"\'\w\ \-\)\/]{0,64}/
+    $os_popen = /os.spopen\([\"\'\w\ \-\)\/]{0,64}/
+    $subprocess = /subprocess.\w{0,32}\([\"\'\/\w\ \-\)]{0,64}/
   condition:
-    $setup and ($setuptools or $distutils) and any of ($s_*)
+    pythonSetup and any of them
 }
 
 rule setuptools_eval : critical {
   meta:
     description = "Python library installer that evaluates arbitrary code"
-    hash_2022_2022_requests_3_0_0_setup = "15507092967fbd28ccb833d98c2ee49da09e7c79fd41759cd6f783672fe1c5cc"
-    hash_2023_requet_2_28_1_setup = "9438107245ebfba792dfa95f7d551392831c20adbcac7d3176797f0f00683ab0"
-    hash_2023_zproxy_1_0_setup = "f3d7eec1ae2eba61715fd0652fa333acc2e4c0d517579392043880aa2f158b62"
   strings:
-    $setup = "setup(" fullword
-    $setuptools = "setuptools"
-    $distutils = "distutils"
-    $s_sys_val = /eval\([\"\'\w\ \-\)\/]{0,64}/ fullword
-    $s_subprocess_val = /exec\([\"\'\/\w\ \-\)]{0,64}/ fullword
+    $f_sys_val = /eval\([\"\'\w\ \-\)\/]{0,64}/ fullword
+    $f_subprocess_val = /exec\([\"\'\/\w\ \-\)]{0,64}/ fullword
+  condition:
+    pythonSetup and any of ($f*)
+}
+
+rule setuptools_b64decode : suspicious {
+  meta:
+    description = "Python library installer that does base64 decoding"
+  strings:
+    $base64 = "b64decode"
   condition:
-    $setup and ($setuptools or $distutils) and any of ($s_*)
+    pythonSetup and any of them
 }
 
-rule setuptools_url_access : high {
+rule setuptools_exec_powershell : critical {
   meta:
-    description = "Python library installer that accesses external URLs"
-    hash_2022_laysound_4_5_2_setup = "4465bbf91efedb996c80c773494295ae3bff27c0fff139c6aefdb9efbdf7d078"
-    hash_2022_2022_requests_3_0_0_setup = "15507092967fbd28ccb833d98c2ee49da09e7c79fd41759cd6f783672fe1c5cc"
-    hash_2022_selenuim_4_4_2_setup = "5c5e1d934dbcbb635f84b443bc885c9ba347babc851cd225d2e18eadc111ecf0"
+    description = "Python library installer that runs powershell"
   strings:
-    $setup = "setup(" fullword
-    $setuptools = "setuptools"
-    $distutils = "distutils"
-    $s_requests = /requests.get\([\"\'\w\ \-\)\/]{0,64}/
-    $s_urlopen = /urlopen\([\"\'\w\ \-\)\/]{0,64}/
+    $powershell = "powershell" fullword
+    $encoded = "-EncodedCommand" fullword
+    $window = "WindowStyle Hidden" fullword
   condition:
-    $setup and ($setuptools or $distutils) and any of ($s_*)
+    setuptools_cmd_exec and any of them
 }
 
-rule setuptools_random : critical {
+rule setuptools_os_path_exists : notable {
   meta:
-    description = "Python library installer that exhibits random behavior"
-    hash_2023_yfinancce_0_1_setup = "3bde1e9207dd331806bf58926d842e2d0f6a82424abd38a8b708e9f4e3e12049"
-    hash_2023_yvper_0_1_setup = "b765244c1f8a11ee73d1e74927b8ad61718a65949e0b8d8cbc04e5d84dccaf96"
+    description = "Python library installer that checks for file existence"
   strings:
-    $setup = "setup(" fullword
-    $setuptools = "setuptools"
-    $distutils = "distutils"
-    $s_sys_val = "import random" fullword
+    $ref = /[\w\.]{0,8}path.exists\([\"\'\w\ \-\)\/]{0,32}/
   condition:
-    $setup and ($setuptools or $distutils) and any of ($s_*)
+    pythonSetup and $ref
 }
 
-rule setuptools_builtins : medium {
+rule setuptools_excessive_bitwise_math : critical {
   meta:
-    description = "Python library installer that directly references builtins"
-    hash_2023_yfinancce_0_1_setup = "3bde1e9207dd331806bf58926d842e2d0f6a82424abd38a8b708e9f4e3e12049"
-    hash_2023_yvper_0_1_setup = "b765244c1f8a11ee73d1e74927b8ad61718a65949e0b8d8cbc04e5d84dccaf96"
+    description = "Python library installer that makes heavy use of bitwise math"
   strings:
-    $setup = "setup(" fullword
-    $setuptools = "setuptools"
-    $distutils = "distutils"
-    $s_sys_val = "__builtins__" fullword
+    $x = /\-{0,1}\d{1,8} \<\< \-{0,1}\d{1,8}/
   condition:
-    $setup and ($setuptools or $distutils) and any of ($s_*)
+    pythonSetup and #x > 20
 }
diff --git a/rules/combo/dropper/python.yara b/rules/combo/dropper/python.yara
@@ -1,40 +1,62 @@
+private rule py_fetcher {
+	meta:
+		description = "fetches content"
+	strings:
+		$http_requests = "requests.get" fullword
+		$http_requests_post = "requests.post" fullword
+		$http_urrlib = "urllib.request" fullword
+		$http_urlopen = "urlopen" fullword
+	condition:
+		any of ($http*)
+}
+
+private rule py_runner {
+	meta:
+		description = "runs programs"
+	strings:
+		$os_system = /os.system\([\"\'\w\ \-\)\/]{0,64}/
+		$os_popen = /os.spopen\([\"\'\w\ \-\)\/]{0,64}/
+		$subprocess = /subprocess.\w{0,32}\([\"\'\/\w\ \-\)]{0,64}/
+	condition:
+		any of them
+}
 
-rule http_open_write_system : high {
+rule py_dropper : suspicious {
   meta:
-    description = "fetch and execute programs"
-    hash_2022_laysound_4_5_2_setup = "4465bbf91efedb996c80c773494295ae3bff27c0fff139c6aefdb9efbdf7d078"
-    hash_2023_JokerSpy_shared = "5fe1790667ee5085e73b054566d548eb4473c20cf962368dd53ba776e9642272"
-    hash_2023_JokerSpy_shared = "39bbc16028fd46bf4ddad49c21439504d3f6f42cccbd30945a2d2fdb4ce393a4"
+  	description = "fetch, stores, and execute programs"
   strings:
-    $http_requests_get = "requests.get" fullword
-    $http_requests_post = "requests.post" fullword
-    $http_urllib = "urllib.request" fullword
-    $http_urlopen = "urlopen" fullword
-    $open = "open("
-    $write = "write("
-    $system = "os.system" fullword
-    $sys_popen = "os.popen" fullword
-    $sys_sub = "subprocess" fullword
+	$open = "open("
+	$write = "write("
   condition:
-    filesize < 16384 and any of ($h*) and $open and $write and any of ($sys*)
+    filesize < 16384 and $open and $write and py_fetcher and py_runner
+}
+
+private rule pythonSetup {
+	strings:
+		$i_distutils = "from distutils.core import setup"
+		$i_setuptools = "setuptools"
+		$setup = "setup("
+	condition:
+		filesize < 2MB and $setup and any of ($i*)
+}
+
+rule setuptools_fetcher : suspicious {
+	meta:
+		description = "setuptools script that fetches content"
+	condition:
+		pythonSetup and py_fetcher
+}
+
+rule setuptools_fetch_run : critical {
+	meta:
+		description = "setuptools script that fetches and executes"
+	condition:
+		setuptools_fetcher and py_runner
 }
 
 rule setuptools_dropper : critical {
-  meta:
-    description = "setuptools script that fetches and executes"
-    hash_2022_laysound_4_5_2_setup = "4465bbf91efedb996c80c773494295ae3bff27c0fff139c6aefdb9efbdf7d078"
-    hash_2022_2022_requests_3_0_0_setup = "15507092967fbd28ccb833d98c2ee49da09e7c79fd41759cd6f783672fe1c5cc"
-    hash_2022_selenuim_4_4_2_setup = "5c5e1d934dbcbb635f84b443bc885c9ba347babc851cd225d2e18eadc111ecf0"
-  strings:
-    $setup = "setup("
-    $setuptools = "setuptools" fullword
-    $http_requests = "requests.get" fullword
-    $http_requests_post = "requests.post" fullword
-    $http_urrlib = "urllib.request" fullword
-    $http_urlopen = "urlopen" fullword
-    $system = "os.system" fullword
-    $sys_popen = "os.popen" fullword
-    $sys_sub = "subprocess" fullword
-  condition:
-    all of ($setup*) and any of ($http*) and any of ($sys*)
+	meta:
+		description = "setuptools script that fetches, stores, and executes"
+	condition:
+		pythonSetup and py_dropper
 }
diff --git a/rules/evasion/bitwise_math.yara b/rules/evasion/bitwise_math.yara
@@ -1,11 +1,24 @@
 
-rule excessive_bitwise_math : medium {
+rule large_bitwise_math : medium {
   meta:
-    description = "excessive use of bitwise math"
+    description = "large amounts of bitwise math"
     hash_2023_yfinancce_0_1_setup = "3bde1e9207dd331806bf58926d842e2d0f6a82424abd38a8b708e9f4e3e12049"
     hash_2023_yvper_0_1_setup = "b765244c1f8a11ee73d1e74927b8ad61718a65949e0b8d8cbc04e5d84dccaf96"
+    hash_2023_aiohttpp_0_1_setup = "cfa4137756f7e8243e7c7edc7cb0b431a2f4c9fa401f2570f1b960dbc86ca7c6"
   strings:
     $x = /\-{0,1}\d{1,8} \<\< \-{0,1}\d{1,8}/
   condition:
     filesize < 128000 and #x > 10
 }
+
+rule excessive_bitwise_math : high {
+  meta:
+    description = "excessive use of bitwise math"
+    hash_2023_yfinancce_0_1_setup = "3bde1e9207dd331806bf58926d842e2d0f6a82424abd38a8b708e9f4e3e12049"
+    hash_2023_yvper_0_1_setup = "b765244c1f8a11ee73d1e74927b8ad61718a65949e0b8d8cbc04e5d84dccaf96"
+    hash_2023_aiohttpp_0_1_setup = "cfa4137756f7e8243e7c7edc7cb0b431a2f4c9fa401f2570f1b960dbc86ca7c6"
+  strings:
+    $x = /\-{0,1}\d{1,8} \<\< \-{0,1}\d{1,8}/
+  condition:
+    filesize < 128000 and #x > 20
+}
diff --git a/rules/evasion/mask_exceptions.yara b/rules/evasion/mask_exceptions.yara
@@ -1,28 +1,30 @@
 import "math"
 
-rule setuptools_no_fail : high {
+private rule pythonSetup {
+  strings:
+    $i_distutils = "from distutils.core import setup"
+    $i_setuptools = "setuptools"
+    $setup = "setup("
+  condition:
+    filesize < 2097152 and $setup and any of ($i*)
+}
+
+rule py_no_fail : notable {
   meta:
-    description = "Python library installer that hides exceptions"
+    description = "Python code that hides exceptions"
     hash_2023_grandmask_3_13_setup = "8835778f9e75e6493693fc6163477ec94aba723c091393a30d7e7b9eed4f5a54"
+    hash_2023_libgrandrandomintel_3_58_setup = "cd211e0f8d84100b1b4c1655e913f40a76beaacc482e751e3a7c7ed126fe1a90"
     hash_2023_py_guigrand_4_67_setup = "4cb4b9fcce78237f0ef025d1ffda8ca8bc79bf8d4c199e4bfc6eff84ce9ce554"
-    hash_2023_py_killtoolad_3_65_setup = "64ec7b05442356293e903afe028637d821bad4444c4e1e11b73a4ff540fe480b"
   strings:
-    $setup = "setup(" fullword
-    $setuptools = "setuptools"
-    $distutils = "distutils"
-    $e_val = /except:.{0,4}pass/ fullword
+    $e_short = /except:.{0,4}pass/ fullword
+    $e_long = /except Exception as.{0,8}pass/ fullword
   condition:
-    $setup and ($setuptools or $distutils) and $e_val
+    any of them
 }
 
-rule setuptools_no_fail2 : high {
+rule setuptools_no_fail : suspicious {
   meta:
     description = "Python library installer that hides exceptions"
-  strings:
-    $setup = "setup(" fullword
-    $setuptools = "setuptools"
-    $distutils = "distutils"
-    $e_val = /except Exception as.{0,8}pass/ fullword
   condition:
-    $setup and ($setuptools or $distutils) and $e_val
+    pythonSetup and py_no_fail
 }
diff --git a/rules/evasion/py_builtins.yara b/rules/evasion/py_builtins.yara
@@ -1,11 +1,37 @@
 
-rule indirect_python_builtins : high {
+rule py_builtins {
+  meta:
+    description = "references Python builtins"
+  strings:
+    $ref = "__builtins__" fullword
+  condition:
+    $ref
+}
+
+rule py_indirect_builtins : suspicious {
   meta:
     description = "Indirectly refers to Python builtins"
     hash_2023_yfinancce_0_1_setup = "3bde1e9207dd331806bf58926d842e2d0f6a82424abd38a8b708e9f4e3e12049"
     hash_2023_yvper_0_1_setup = "b765244c1f8a11ee73d1e74927b8ad61718a65949e0b8d8cbc04e5d84dccaf96"
+    hash_2023_aiohttpp_0_1_setup = "cfa4137756f7e8243e7c7edc7cb0b431a2f4c9fa401f2570f1b960dbc86ca7c6"
   strings:
     $val = /getattr\(__builtins__,[ \w\.\)\)]{0,64}/
   condition:
     any of them
 }
+
+private rule pythonSetup {
+  strings:
+    $i_distutils = "from distutils.core import setup"
+    $i_setuptools = "setuptools"
+    $setup = "setup("
+  condition:
+    filesize < 2097152 and $setup and any of ($i*)
+}
+
+rule setuptools_builtins : notable {
+  meta:
+    description = "Python library installer that references builtins"
+  condition:
+    pythonSetup and py_builtins
+}
diff --git a/rules/evasion/py_setuptools-random.yara b/rules/evasion/py_setuptools-random.yara
@@ -0,0 +1,19 @@
+import "math"
+
+private rule pythonSetup {
+	strings:
+		$i_distutils = "from distutils.core import setup"
+		$i_setuptools = "setuptools"
+		$setup = "setup("
+	condition:
+		filesize < 2MB and $setup and any of ($i*)
+}
+
+rule setuptools_random : critical {
+	meta:
+		description = "Python library installer that exhibits random behavior"
+	strings:
+		$ref = "import random"
+	condition:
+		pythonSetup and $ref
+}
diff --git a/rules/exec/program.yara b/rules/exec/program.yara
@@ -50,18 +50,33 @@ rule exec_cmd_run : medium {
     any of them
 }
 
-rule perl_system : medium {
+rule perl_system : notable {
   meta:
     syscall = "execve"
     pledge = "exec"
     description = "executes external programs"
-    hash_2023_0xShell_0xShellori = "506e12e4ce1359ffab46038c4bf83d3ab443b7c5db0d5c8f3ad05340cb09c38e"
-    hash_2023_0xShell_root = "3baa3bfaa6ed78e853828f147c3747d818590faee5eecef67748209dd3d92afb"
-    hash_2023_0xShell_untitled = "39b2fd6b4b2c11a9cbfc8efbb09fc14d502cde1344f52e1269228fc95b938621"
   strings:
-    $ref = "system("
+    $system = /system\([\"\'\w\ \-\)\/]{0,64}/
+    $perl = "perl" fullword
   condition:
-    all of them
+    filesize < 65535 and $perl and $system
+}
+
+rule py_subprocess : notable {
+  meta:
+    syscall = "execve"
+    pledge = "exec"
+    description = "execute external program"
+    ref = "https://man7.org/linux/man-pages/man2/execve.2.html"
+    hash_2022_2022_requests_3_0_0_setup = "15507092967fbd28ccb833d98c2ee49da09e7c79fd41759cd6f783672fe1c5cc"
+    hash_2023_grandmask_3_13_setup = "8835778f9e75e6493693fc6163477ec94aba723c091393a30d7e7b9eed4f5a54"
+    hash_2023_libgrandrandomintel_3_58_setup = "cd211e0f8d84100b1b4c1655e913f40a76beaacc482e751e3a7c7ed126fe1a90"
+  strings:
+    $naked = "subprocess"
+    $val = /subprocess\.\w{1,16}[\(\"\/\w\'\.\- \,\[\]]{0,64}/
+    $os_system = /os.system\([\"\'\w\ \-\)\/]{0,64}/
+  condition:
+    any of them
 }
 
 rule subprocess : medium {