From 5f06e60fddd69a305e5da57f9aef6351420be830 Mon Sep 17 00:00:00 2001 From: = <=> Date: Mon, 25 Aug 2025 12:35:27 -0400 Subject: [PATCH] New Master candidate --- AirlockTools.py | 44 +++++++++++++++++++------- hashtest.py | 5 --- utils/allowlist.py | 70 +----------------------------------------- utils/hashfunctions.py | 36 ++++++++++++++-------- utils/pathfunctions.py | 30 ++---------------- 5 files changed, 60 insertions(+), 125 deletions(-) delete mode 100644 hashtest.py diff --git a/AirlockTools.py b/AirlockTools.py index 00c5d69..7c1060a 100644 --- a/AirlockTools.py +++ b/AirlockTools.py @@ -19,11 +19,11 @@ import utils.getdeviceevents import utils.allowlist import utils.hashfunctions import utils.pathfunctions +import utils.allowfunctions import utils.colortext as ct import urllib3 import pandas as pd - urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning) dotenv.load_dotenv() @@ -121,7 +121,7 @@ def menu_prepare_to_enforce(): #If the directorys where we're going to store our output dont exist, make them. if not os.path.exists("dataframe_html"): os.makedirs("dataframe_html") if not os.path.exists("dataframe_csv"): os.makedirs("dataframe_csv") - if not os.path.exists("approvals"): os.makedirs("approvals") + if not os.path.exists("manuallyapproved"): os.makedirs("manuallyapproved") df_aggregated_combo = pd.DataFrame() while True: @@ -177,14 +177,25 @@ def menu_prepare_to_enforce(): print(ct.colorText(" [✓] This step has been completed","green")) else: print(ct.colorText(" [✗] This step has not been completed","red")) - + + print(ct.colorText(f"7. Manually review the files \\dataframe_csv\\df_hashes_needing_approval_{first_policy}_{second_policy}.csv and dataframe_csv\\df_automatically_approved_hashes_{first_policy}_{second_policy}.csv", "cyan")) + print(ct.colorText(" Remove the rows containing hashes you do not approve of, and those you would not approve of without metarules.", "cyan")) + print(ct.colorText(" If metarules need to be created, please make note of them, and remove the row from the csv.", "cyan")) + print(ct.colorText(" When complete, move both csv files to the directory 'manuallyapproved' and choose this option to combine these approved hashes with the automatically approved hashes", "cyan")) + + if os.path.isfile(f"dataframe_csv\\df_all_approved_hashes_{first_policy}_{second_policy}.csv"): + print(ct.colorText(" [✓] This step has been completed","green")) + else: + print(ct.colorText(" [✗] This step has not been completed","red")) + + """ print(ct.colorText("7. Compare potential path exclusions with allowed hashes", "cyan")) if os.path.exists(f"dataframe_csv\\df_allowed_paths_{first_policy}_{second_policy}.csv") == True: print(ct.colorText(" [✓] This step has been completed","green")) else: print(ct.colorText(" [✗] This step has not been completed","red")) - + """ print(ct.colorText("Q. Quit", "cyan")) @@ -268,6 +279,21 @@ def menu_prepare_to_enforce(): else: print(ct.colorText(f"Please Augment your data with hash threat info using step 4 prior to attempting this step","red")) + elif choice == "7": + if os.path.isfile(f"manuallyapproved\\df_hashes_needing_approval_{first_policy}_{second_policy}.csv"): + df1 = tryToReadCSV(f"manuallyapproved\\df_hashes_needing_approval_{first_policy}_{second_policy}.csv") + df2 = tryToReadCSV(f"manuallyapproved\\df_automatically_approved_hashes_{first_policy}_{second_policy}.csv") + df_all_approved_hashes = pd.concat([df1 , df2], ignore_index=True) + df_aggregated_combo.to_html(f"dataframe_html\\df_all_approved_hashes_{first_policy}_{second_policy}.html", index=False) + df_aggregated_combo.to_csv(f"dataframe_csv\\df_all_approved_hashes_{first_policy}_{second_policy}.csv", index=False) + + + elif choice == "Q": + break + else: + print(ct.colorText("Invalid choice. Please try again.", "red")) + + """ elif choice == "7": if os.path.exists(f"dataframe_csv\\df_hashes_needing_approval_{first_policy}_{second_policy}.csv") and os.path.exists(f"dataframe_csv\\df_automatically_approved_hashes_{first_policy}_{second_policy}.csv") and os.path.exists(f"dataframe_csv\\df_unapproved_hashes__{first_policy}_{second_policy}.csv"): allowpaths = utils.allowfunctions.filter_and_drop(pd.read_csv(f"dataframe_csv\\df_automatically_approved_hashes_{first_policy}_{second_policy}.csv"),tryToReadCSV(f"dataframe_csv\\df_path_eligible_{first_policy}_{second_policy}.csv"), path_exclusion_constant) @@ -276,13 +302,9 @@ def menu_prepare_to_enforce(): print(ct.colorText(f"Allowable paths determined","green")) else: print(ct.colorText(f"Please complete step 6 prior to attempting this step","red")) - - elif choice == "Q": - break - - else: - print(ct.colorText("Invalid choice. Please try again.", "red")) - + """ + + def tryToReadCSV(csv): try: df =pd.read_csv(csv) diff --git a/hashtest.py b/hashtest.py deleted file mode 100644 index b0dab88..0000000 --- a/hashtest.py +++ /dev/null @@ -1,5 +0,0 @@ -hashes = '' -while True: - inputhash = input("Hash: ") - hashes = hashes + ',' + inputhash - print(hashes) \ No newline at end of file diff --git a/utils/allowlist.py b/utils/allowlist.py index 0528be7..6cc6e2d 100644 --- a/utils/allowlist.py +++ b/utils/allowlist.py @@ -19,7 +19,6 @@ import os import time import utils.colortext as ct - def pullPolicyExechistories(url, choice, policiesnames, outputjson: bool): headers = { @@ -74,71 +73,4 @@ def listPolicies(url): policyids.append(list['groupid']) choice = input(ct.colorText("Select Policy Group: ", "white")) choice = int(choice) - 1 - checkpoint = '000000000000000000000000' - json_output = {'error': 'Success', 'response': {'exechistories': []}} - while True: - json_response_data = checkpoint_stomper(checkpoint, url, policiesnames[choice], headers) - if not json_response_data['response']['exechistories']: - break - for index, item in enumerate(json_response_data['response']['exechistories']): - if index == len(json_response_data['response']['exechistories']) -1: - checkpoint = item['checkpoint'] - print(f"Date Greater than 30 Days, Stepping to new Checkpoint. {item['checkpoint']}") - else: - if (datetime.date.today() - datetime.timedelta(days=10) > datetime.datetime.strptime(item['datetime'].replace(' +0000 UTC', ''), '%Y-%m-%dT%H:%M:%SZ').date()): - pass - else: - #json_output['response']['exechistories'].append(json_response_data['response']['exechistories'][1]) - for output in json_response_data['response']['exechistories']: - json_output['response']['exechistories'].append(output) - json_output = json.dumps(json_output) - if outputjson == True: - return json_output - #endpoint = url + '/v1/logging/exechistories' - #payload_dict = { - # "type":[1, 2, 6, 7], - # "checkpoint":"000000000000000000000000", - # "policy": [policiesnames[choice]] - #} - #payload = json.dumps(payload_dict) - #print(payload) - #response = requests.request("POST", endpoint, headers=headers, data=payload, verify=False) - #parse_text = json.loads(response.text) - #text_response = checkpoint_stomper(parse_text['response']['exechistories'], url, policiesnames[choice]) -# - #if outputjson == False: - # return response - # - #parse_text = json.loads(response.text) - # - #for item in parse_text['response']['exechistories']: - # print(item['checkpoint']) - # print(item['datetime']) - # print(item['hostname']) - # print(item['filename']) - # checkpoint_stomper(item['checkpoint'], endpoint, headers, policiesnames[choice]) - -def checkpoint_stomper(checkpoint, url, policy, headers): - endpoint = url + '/v1/logging/exechistories' - payload_dict = { - "type":[1,2,6,7], - "checkpoint": checkpoint, - "policy": [policy] - } - payload = json.dumps(payload_dict) - response = requests.request("POST", endpoint, headers=headers, data=payload, verify=False) - parse_text = json.loads(response.text) - return parse_text - #for index, item in enumerate(parse_text): - # if index == len(parse_text) - 1: - # checkpoint = item['checkpoint'] - # print(f"Time: {item['datetime']} Checkpoint: {item['checkpoint']}") - # response_fuzzer(checkpoint, url, policyname) - # else: - # if (datetime.date.today() - datetime.timedelta(days=30) > datetime.datetime.strptime(item['datetime'].replace( ' +0000 UTC', ''), '%Y-%m-%dT%H:%M:%SZ').date()): - # pass - # else: - # response_fuzzer(checkpoint, url, policyname) - - - print("Finished") + return choice, policiesnames \ No newline at end of file diff --git a/utils/hashfunctions.py b/utils/hashfunctions.py index 9b9c59b..c193284 100644 --- a/utils/hashfunctions.py +++ b/utils/hashfunctions.py @@ -90,28 +90,38 @@ def categorizeHashes(aug_df: pd.DataFrame, threat_tolerance: int, untrusted_publ df = aug_df.copy() - def reputationtool(row, threat_tolerance): - if row["reputation_scannermatch"] == "N/A": - return True + def reputationtool(row): + val = row["reputation_scannermatch"] + if pd.isna(val) or val == "N/A": + return row["publisher_y"] == "Not Signed" try: return int(val) > threat_tolerance except (ValueError, TypeError): return row["publisher_y"] == "Not Signed" - mask_needsreview = (df["publisher_y"] == "Not Signed") & df.apply(lambda row: reputationtool(row, threat_tolerance), axis=1) + df["reputation_flag"] = df.apply(reputationtool, axis=1) - mask_approved = ( - ((df["publisher_y"] != "Not Signed") & (~df["publisher_y"].isin(untrusted_publishers))) & - (df["publisher_y"] != "Not Signed") # explicitly signed - ) | ( - (df["publisher_y"] == "Not Signed") & - (~df.apply(lambda row: reputationtool(row, threat_tolerance), axis=1)) & - (~df["publisher_y"].isin(untrusted_publishers)) # exclude untrusted even if unsigned + mask_needsreview = ( + ((df["publisher_y"] == "Not Signed") & df["reputation_flag"]) | + (df["reputation_status"] == "UNKNOWN") ) + mask_approved = ( + ( + (df["publisher_y"] != "Not Signed") & + ~df["publisher_y"].isin(untrusted_publishers) & + ~df["reputation_status"].isna() + ) | + ( + (df["publisher_y"] == "Not Signed") & + ~df["reputation_flag"] & + ~df["publisher_y"].isin(untrusted_publishers) & + ~df["reputation_status"].isna() + ) + ) needsreview_df = df[mask_needsreview] approved_df = df[mask_approved] - remaining_df = df[~(mask_needsreview | mask_approved)] + unapproved_df = df[~(mask_needsreview | mask_approved)] - return needsreview_df, approved_df, remaining_df + return needsreview_df, approved_df, unapproved_df diff --git a/utils/pathfunctions.py b/utils/pathfunctions.py index 72bdae2..ab8e166 100644 --- a/utils/pathfunctions.py +++ b/utils/pathfunctions.py @@ -57,45 +57,21 @@ def filepathInitialGroup(df: pd.DataFrame): break return join_parts(prefix) - # Step 6: Group directories by shared prefix using custom logic - """ - Loop through each directory path - directories: list of all directory paths. - groups: will hold lists of grouped directories. - used: tracks which directories have already been grouped. - """ + # Step 6: Group directories by shared prefix directories = df["directory"].tolist() groups = [] used = set() - #For Each directory, compare it with others - """ - Skip if already grouped. - Start a new group with the current path. - parts_i is the list of folder names in the path (e.g., ["C:", "Users", "John", "Documents"]). - """ - for i, path in enumerate(directories): if path in used: continue group = [path] parts_i = get_parts(path) - #Compare with all other directories: For each other directory, split it into parts and find the common prefix (shared folder structure). - """ - Logic: - If the directory is deep (>3 parts) and shares at least 3 parts → group it. - If it's exactly 3 parts long and shares at least 2 → group it. - Or, if it shares all but one part and is deep → group it. - These rules are designed to: - Group directories that are closely related in structure. - Avoid grouping unrelated paths that just happen to start similarly. - """ - for j in range(i + 1, len(directories)): parts_j = get_parts(directories[j]) common = os.path.commonprefix([parts_i, parts_j]) - #Apply grouping rules + if (len(parts_i) > 3 and len(common) >= 3) or (len(parts_i) == 3 and len(common) >= 2): group.append(directories[j]) used.add(directories[j]) @@ -123,7 +99,7 @@ def filepathInitialGroup(df: pd.DataFrame): path_eligible = grouped_df[grouped_df["depth"] > 2].drop(columns=["depth"]) path_ineligible = grouped_df[grouped_df["depth"] <= 2].drop(columns=["depth"]) - # Step 10: Move entries from eligible to ineligible if grouped_directory contains 'C:\Users' or 'c$\Users' + # Step 10: Move entries from eligible to ineligible if grouped_directory contains excluded directories mask = path_eligible["grouped_directory"].str.contains(r"(?i)(?:\\Users|\\c\$\\Users|inetpub\\wwwroot|windows\\temp)", na=False) move_to_ineligible = path_eligible[mask] path_eligible = path_eligible[~mask]