alexgreensh/repo-forensics / skills/repo-forensics
repo-forensics
FlaggedSecurity forensics for git repos, AI skills, and MCP servers. Audits dependencies, detects prompt injection, credential theft, runtime dynamism, manifest drift, known CVEs, CISA KEV (actively exploited) vulns, and 2026 attack patterns. Not for fixing vulnerabilities or pentesting.
data/yara/manifest.json
View on GitHub{
"schema_version": "1.0",
"pack": "yara",
"pack_version": 1,
"generated": "2026-08-04",
"files": [
{
"path": "malware.yar",
"sha256": "a6c773232622e6b4bd1b236c9bf0e8ac453cd406f7ecaf07a16f25129a3f1300"
},
{
"path": "webshells.yar",
"sha256": "b3157112732722063e56819e4c3e73d298edb245a8a284474e69a258847d005f"
},
{
"path": "cryptominers.yar",
"sha256": "0764524179a3dcebad0ee2f2db95bc168ecc8941a079876a5655d2edc4b99bb9"
},
{
"path": "hacktools.yar",
"sha256": "e8fadf1a37140d05da6ba9bb0eafb91d83ef1cc024d4a69ab84792f21d701dbb"
}
],
"rule_meta": {
"Webshell_PHP_Eval_Inflate": {
"id": "YR-WEB-001",
"severity": "critical",
"confidence": 0.95,
"category": "webshell",
"title": "PHP webshell (eval/assert + gzinflate + base64 chain)",
"explanation": "Conjunctive match of an executor (eval( or assert() with a decode primitive (gzinflate/base64_decode) is the canonical PHP obfuscated-webshell chain. v2.13.1 reclaims assert() as an alternative executor (the prior `not $assert` exclusion existed only to spare legit assert(decode(...)) unit-test code, which the context gate now demotes). Multi-string conjunction makes this a confirmed family indicator, not a single-token guess.",
"attacker": "Operator with upload access planting a webshell",
"boundary": "Web server process executes attacker PHP",
"asset": "Host execution + data on the web server"
},
"Webshell_PHP_Exec_Passthru": {
"id": "YR-WEB-002",
"severity": "critical",
"confidence": 0.9,
"category": "webshell",
"title": "PHP webshell (system/passthru/shell_exec/exec/popen exec chain)",
"explanation": "Two command-execution primitives (system/passthru/shell_exec/proc_open/exec/popen) co-occurring with a request input binding ($_REQUEST/$_GET/$_POST) is the shape of a PHP command-execution webshell. v2.13.1 widens the exec set (exec(/popen() and the input set ($_POST); $_POST in PHP-tutorials/tests is absorbed by the context gate (prose-doc/test-fixture -> inferred).",
"attacker": "Operator with upload access planting a webshell",
"boundary": "Web server process executes attacker PHP",
"asset": "Host execution + data on the web server"
},
"Webshell_JSP_Runtime_Exec": {
"id": "YR-WEB-003",
"severity": "critical",
"confidence": 0.92,
"category": "webshell",
"title": "JSP webshell (Runtime.exec + request.getParameter chain)",
"explanation": "Runtime.getRuntime().exec combined with request.getParameter is the canonical JSP command-execution webshell pattern.",
"attacker": "Operator with upload access planting a webshell",
"boundary": "Web container executes attacker JSP",
"asset": "Host execution + data on the application server"
},
"Malware_Reverse_Shell_Stager": {
"id": "YR-MAL-001",
"severity": "high",
"confidence": 0.85,
"category": "malware",
"title": "Reverse shell stager (socket + dup2 + shell spawn)",
"explanation": "socket.socket + dup2 + a shell binary (/bin/sh or cmd.exe) is the classic Python reverse-shell stager. The conjunction is a confirmed indicator, not a benign network client.",
"attacker": "Malware operator establishing a callback",
"boundary": "Host process opens outbound socket to attacker",
"asset": "Interactive shell access on the host"
},
"Malware_Powershell_Encoded_Stager": {
"id": "YR-MAL-002",
"severity": "high",
"confidence": 0.85,
"category": "malware",
"title": "PowerShell encoded-command stager",
"explanation": "powershell with an encoded-command flag (-EncodedCommand or the short -enc/-ec/-e) paired with a stealth flag (-WindowStyle Hidden / -NoProfile or the shorthand -w hidden / -nop) is the standard dropper/stager invocation. v2.13.1 accepts the short-flag encodings and shorthand stealth flags real droppers use; PS droppers documented in prose/tutorials are demoted by the context gate. EncodedCommand hides the payload from plain-text inspection.",
"attacker": "Malware operator invoking an obfuscated stager",
"boundary": "User-context PowerShell executes attacker payload",
"asset": "Host execution under the user context"
},
"Malware_HTA_Dropper": {
"id": "YR-MAL-003",
"severity": "high",
"confidence": 0.8,
"category": "malware",
"title": "HTA dropper (mshta + WScript.Shell.Run)",
"explanation": "mshta paired with WScript.Shell and a Run/Shell.Application invocation is the shape of an HTA dropper executing a remote or local payload.",
"attacker": "Malware operator delivering an HTA dropper",
"boundary": "User-context script host executes attacker payload",
"asset": "Host execution under the user context"
},
"Cryptominer_XMRig_Stratum": {
"id": "YR-CRY-001",
"severity": "high",
"confidence": 0.9,
"category": "cryptominer",
"title": "XMRig / stratum cryptominer pool indicator",
"explanation": "A stratum pool URI (stratum+tcp:// or stratum+ssl://) co-occurring with an XMRig/cryptonight/randomx marker (incl. the quoted \"rx/0\" config form and common public Monero pool hosts: pool.supportxmr.com/xmrpool./nanopool./moneroocean/hashvault) is a confirmed cryptominer pool indicator. v2.13.1 widens the pool/algorithm marker set; mining-ops docs and pool lists in blocklists are demoted by the context gate.",
"attacker": "Malware operator hijacking CPU for mining",
"boundary": "Host process connects to a mining pool",
"asset": "CPU cycles + electricity on the host"
},
"Cryptominer_Coinhive_Browser": {
"id": "YR-CRY-002",
"severity": "medium",
"confidence": 0.8,
"category": "cryptominer",
"title": "In-browser cryptominer (CoinHive / WASM wrapper)",
"explanation": "Two CoinHive/WASM-wrapper markers co-occurring indicate an in-browser cryptominer payload. Medium severity: browser-context, lower host impact than a native miner.",
"attacker": "Site operator hijacking visitor CPU for mining",
"boundary": "Visitor browser executes miner script",
"asset": "Visitor CPU cycles"
},
"Hacktool_Mimikatz": {
"id": "YR-HCK-001",
"severity": "medium",
"confidence": 0.8,
"category": "hacktool",
"title": "Mimikatz credential-dumper indicator",
"explanation": "Two Mimikatz markers (sekurlsa/logonpasswords/mimikatz/gentilkiwi) co-occurring indicate the Mimikatz credential-dumping tool. Medium: a hacktool presence, not an active exploit by itself.",
"attacker": "Operator running a credential-dumper",
"boundary": "Operator process reads LSASS credential material",
"asset": "Cached credentials on the host"
},
"Hacktool_Metasploit": {
"id": "YR-HCK-002",
"severity": "medium",
"confidence": 0.75,
"category": "hacktool",
"title": "Metasploit framework indicator",
"explanation": "Two Metasploit markers (metasploit/msfconsole/exploit-multi-handler/msfvenom) co-occurring, OR a meterpreter payload alongside msfvenom, indicate the Metasploit exploitation framework. v2.13.1 adds the meterpreter payload alternative. Medium: framework presence, not an active exploit.",
"attacker": "Operator running an exploitation framework",
"boundary": "Operator process launches exploit modules",
"asset": "Target services on the network"
},
"Hacktool_LSASS_Dump": {
"id": "YR-HCK-003",
"severity": "medium",
"confidence": 0.8,
"category": "hacktool",
"title": "LSASS credential-dump indicator",
"explanation": "lsass.exe paired with a dump primitive (MiniDump/OpenProcess/comsvcs.dll/procdump) is the shape of an LSASS credential-dump. v2.13.1 adds procdump as a dump primitive. Medium: hacktool presence with credential-theft intent.",
"attacker": "Operator dumping LSASS credential material",
"boundary": "Operator process reads LSASS memory",
"asset": "Cached credentials on the host"
}
}
}Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |