Secrets & Credential Leaks
Secrets management and credential leak prevention address one of the most common and impactful security failures in modern software development. Hardcoded API keys, database passwords, cloud credentials, and private keys regularly appear in source code repositories, CI/CD configurations, container images, client-side JavaScript, and log files. Tools like TruffleHog, GitLeaks, and GitHub Secret Scanning detect exposed credentials in repositories, while vault solutions like HashiCorp Vault, AWS Secrets Manager, and cloud KMS services provide secure runtime secret injection. The impact of leaked credentials can be devastating — exposed AWS keys can lead to full cloud account compromise within minutes, and leaked database credentials can result in complete data breaches. Prevention requires secrets scanning in CI/CD pipelines, pre-commit hooks, environment-based secret injection, and credential rotation policies.
Credentials leak through history, not through the current commit
The awkward property of a leaked secret is that removing it does not unleak it. A key committed and deleted in the next commit is still in the git history, still in every clone, still in forks, and still in whatever mirrors and caches exist — and for a public repository, quite possibly in a dataset somebody built by scraping. Detection at the current HEAD finds almost nothing; scanning history is what finds the real exposure, and rotation, not deletion, is what actually remediates it.
The places secrets accumulate are predictable once you look for them. Version control history is first. CI and build logs are second and are frequently overlooked, because a variable that is masked in one job gets echoed by a debugging line in another, and logs are often more widely readable than the repository. Client-side bundles come third, and the recurring pattern is an API key embedded in JavaScript or a mobile package under the belief that minification or compilation is concealment. After that: container images with credentials baked into a layer, backups and database dumps, configuration files in deployment artifacts, Slack and ticket attachments, and cloud metadata reachable through SSRF.
Detection is a precision problem rather than a recall problem. Regex rules for well-known formats work well because providers gave their keys recognizable prefixes; generic high-entropy detection produces enormous false-positive volume and gets ignored, which is worse than not running it. Provider-side verification — checking whether a candidate key is live — is what makes a scanner's output actionable.
The structural fix is to reduce what a leak is worth. Short-lived credentials issued by a broker, workload identity federation instead of long-lived keys in CI, per-service scoping so one credential does not open everything, and monitoring for use from unexpected locations all shrink the blast radius of an exposure you did not catch. Pre-commit hooks and repository scanning reduce frequency; they will not reach zero, and planning for that is the point.
| Date Added | Link | Excerpt |
|---|---|---|
| 2026-11-19 NEW 2026 | Tensei auth bypass via hard-coded JWT secret intermediate AuthN JWT | Tensei versions up to 0.11.5 have a critical authentication bypass vulnerability. This flaw stems from a hard-coded JWT secret, allowing unauthorized access to the system. The vulnerability enables attackers to circumvent security measures and gain access to sensitive information or functionalities. |
| 2026-11-19 NEW 2026 | Java4Shell: When Decoding base64 Leads to RCE intermediate RCE | Java4Shell, addressed by OpenJDK and Oracle on November 17, 2026, enables pre-authentication Remote Code Execution (RCE) in Java applications with default settings. This vulnerability, exploitable in identity, CI/CD, artifact repository, and collaboration tools, poses a significant secret-theft risk due to its presence in applications that store sensitive credentials. → blog.gitguardian.com |
| 2026-11-19 NEW 2026 | CAST Imaging hardcoded Keycloak auth token intermediate AuthN | CAST Imaging's software has a security vulnerability where a hardcoded authentication token for Keycloak was discovered. This token, if exploited, could allow unauthorized access to sensitive information or systems. The hardcoded nature of the token means it's present in the application's code, making it a predictable target for attackers. |
| 2026-11-19 NEW 2026 | Cursor Says Stolen Fireworks AI Key Led to Trojanized Claude Code Extension Incident news AI Supply Chain | Cursor reports that a compromised Fireworks AI key was the root cause of a security incident affecting its Claude Code extension. This breach allowed for data theft and unauthorized access, potentially impacting as many as 5,000 users. The incident highlights the critical need for robust key management to prevent such vulnerabilities. → thehackernews.com |
| 2026-11-19 NEW 2026 | Outstatic GitHub proxy sends OAuth token to non-GitHub origins intermediate SSRF | Outstatic's GitHub REST proxy has a vulnerability where authenticated OAuth tokens could be sent to origins other than GitHub. This means sensitive tokens might be exposed to unintended third parties, potentially allowing unauthorized access to user GitHub accounts and data. |
| 2026-11-19 NEW 2026 | temporal-proxy gRPC listener exposes static credentials intermediate SSRF | A vulnerability in the temporal-proxy gRPC listener has been discovered where static credentials can be exposed to callers sending requests. This means unauthorized parties could potentially gain access to sensitive information or perform actions they shouldn't be able to by exploiting these exposed credentials. The issue lies within the gRPC interface of the temporal-proxy service. |
| 2026-11-18 NEW 2026 | Two malicious agent-framework releases hit PyPI in 33 minutes news Supply Chain | Two malicious agent-framework packages were uploaded to PyPI within 33 minutes, each containing a credential stealer. These packages would execute on every Python startup, attempting to steal secrets from Azure, Azure DevOps, and GitHub Actions. This incident highlights the ongoing threat of supply chain attacks targeting developer environments and cloud-based secrets. → aikido.dev |
| 2026-11-18 NEW 2026 | Farmasimplex shared OAuth credentials news AuthN | Farmasimplex's CeliachiaRL integration component inadvertently shared sensitive OAuth credentials. This vulnerability exposed user data and authentication tokens, allowing unauthorized access to user accounts. The issue stemmed from improperly secured configuration files within the component. Users are advised to update their CeliachiaRL integration to the latest version to patch this security flaw. |
| 2026-11-18 NEW 2026 | N-able N-central predictable credential news AuthN | N-able N-central versions prior to 2026.4.1 HF3 have a security vulnerability where credentials for SQL Server access are predictable. This flaw could allow unauthorized users to gain access to sensitive data stored within N-central. The issue has been addressed in later versions. |
| 2026-11-18 NEW 2026 | ICRTouch PocketTouch hard-coded defaults news | ICRTouch PocketTouch, specifically version 1.700 build 2025.3726, has a security vulnerability related to hard-coded default values. These defaults are not configurable and could potentially be exploited, although details on the exact impact and mitigation are not provided. |
| 2026-11-18 NEW 2026 | I Ignored a NEXT_PUBLIC_ Token for a Week. Then It Unlocked Blind SQLi and the Whole Database intermediate Bug Bounty SQLi | A young bug hunter discovered a blind SQL injection vulnerability by noticing a `NEXT_PUBLIC_` token that had been overlooked for a week. This overlooked token ultimately granted access to the entire database. |
| 2026-11-18 NEW 2026 | Microsoft agent-framework Compromised: Credential Stealer Targets Azure and CI Pipelines news 4 min read AI Python Supply Chain | Library versions 1.26.1 and 1.26.2 of Microsoft's agent-framework package were maliciously published to PyPI, dropping credential-stealing .pth files that run on interpreter startup. These versions target Azure and CI pipelines, collecting environment variables, process environments, and credential files like Azure CLI configs, npmrc, and OIDC tokens for services including Azure DevOps and GitHub Actions. The malware exfiltrates compressed JSON data to a hardcoded webhook.site URL. → wiz.io |
| 2026-11-17 NEW 2026 | Why Is Credential Rotation So Hard? Follow One Leaked Production Key intermediate 13 min read | Tool for analyzing secrets sprawl, this entry walks through remediating a single leaked production API key. It contrasts a manual, multi-day rotation process involving hardcoded fallbacks, config files, pipeline variables, vaults, and Jira tickets with a faster, hours-long remediation. The analysis highlights how GitGuardian streamlines identifying all credential copies, their owners, and vault matches, enabling proactive management and potentially skipping manual rotation entirely, especially in light of the 28.65 million new hardcoded secrets found in public GitHub commits in 2025. → blog.gitguardian.com |
| 2026-11-17 NEW 2026 | The “Don’t Ask Again” Button That Leaked 2,539 Secrets to GitHub intermediate | A seemingly innocuous "Don't Ask Again" button on a local development tool led to the exposure of 2,539 secrets, including database passwords and cloud keys, on GitHub. This UI element, intended to simplify user experience, inadvertently allowed sensitive credentials to be committed and publicly shared. The incident highlights how even minor design choices can have significant security implications, underscoring the need for careful consideration of potential data leakage in user interface development. |
| 2026-11-17 NEW 2026 | WPPConnect Compromised: Malicious Releases Steal Secrets and WhatsApp Sessions Through the Project's Own Pipeline news 10 min read Supply Chain | Library releases of @wppconnect-team/wppconnect 2.4.1 and @wppconnect/server 2.10.82 contained a malicious "diagnostics" module designed to steal host environment variables, .env files, and WhatsApp session tokens, exfiltrating them to a Telegram bot. This payload executed upon import or server startup, and was present in Docker images and GitHub nightly releases. The release tags were later rewritten to remove the malicious code, obscuring the discrepancy between the code on GitHub and the versions available on npm. → stepsecurity.io |
| 2026-11-17 NEW 2026 | New Mior and Skyfall Attacks Let Rogue Cloud Hosts Steal Secrets From AMD SEV-SNP VMs advanced 6 min read | Library detailing attacks Mior (CVE-2026-0462) and Skyfall (CVE-2026-40671), which compromise AMD SEV-SNP protected VMs. Mior leverages hypervisor and BIOS control to modify VM memory and fake attestation reports, while Skyfall exploits Transient Scheduler Attacks (TSA) on failing VM starts to leak data. These attacks target specific AMD EPYC server processors and require firmware and microcode updates from server vendors. → thehackernews.com |
| 2026-11-17 NEW 2026 | My AI can use my API key. It just can’t see it advanced AI | Researchers developed a novel credential proxy for AI agents, allowing them to securely access APIs without directly viewing sensitive credentials. This proxy is embedded within the AI model's weights, acting as an intermediary. The AI can successfully utilize the API key for operations, but its internal architecture prevents it from directly accessing or "seeing" the key itself, enhancing security by minimizing exposure. |
| 2026-11-17 NEW 2026 | Amazon Uncovers AI Agent Campaign Targeting LiteLLM, Dify, and Other Apps for API Keys news 6 min read AI API Sec | Library for detecting AI agent campaigns targeting LLM applications like LiteLLM, Dify, and One API. These agents exploit vulnerabilities such as default credentials, unfinished installations, and flaws in Langflow and RAGFlow to steal AI provider keys. The library aids in identifying indicators like ARTEX worker scripts and the use of tools for scanning GitHub, npm, PyPI, and other sources for exposed secrets and misconfigured cloud services. → thehackernews.com |
| 2026-11-16 NEW 2026 | Who Owns Your Agent’s Credentials? beginner AI | Agent identity is a critical, yet often overlooked, gap in AI security. As AI agents perform tasks like sending emails or accessing customer data, questions of ownership and security surrounding their credentials arise. This lack of attention to agent identity management leaves organizations vulnerable, as the systems and data these agents interact with are often unprotected from unauthorized access or malicious use. Addressing this emerging threat requires a proactive approach to secure and manage AI agent identities. |
| 2026-11-16 NEW 2026 | AI Agent Traces Are a Credential Store: What Cursor’s Breach Shows About Secrets in Agent Logs beginner AI | The Cursor breach highlights a critical security vulnerability: AI agent traces can inadvertently become credential stores. When sensitive information, like API keys or passwords, is fed into an AI agent's context, it can be logged and stored in the agent's traces, datasets, or logs. This creates a significant risk if these logs are compromised. The breach serves as a stark reminder of the importance of rigorously sanitizing AI agent inputs and implementing robust security measures to protect secrets within these systems. |
| 2026-11-16 NEW 2026 | Find leaked secrets in Git and rotate them before they are used intermediate | This article, originally from the Rokkhe blog on Hashnode and AI-assisted, focuses on identifying and rotating leaked secrets within Git repositories. The core idea is to proactively discover sensitive information that may have been accidentally exposed in code history. By finding these secrets before they can be exploited, organizations can then implement remediation steps, such as rotating credentials, to prevent potential security breaches. The emphasis is on secure Git practices and the importance of continuous monitoring for exposed sensitive data. |
| 2026-11-16 NEW 2026 | wfx GitHub CI Secret Stealer and the bee-js npm Compromise news Supply Chain | This report details two security incidents from the MAG212 CYBER Weekly on November 14, 2026. The first incident involves a secret stealer found within wfx GitHub CI. The second incident concerns a compromise of the bee-js npm package. Both issues have been confirmed by the respective vendors and reported by researchers. Further details on the nature of the exploits and their impact are provided. |
| 2026-11-16 NEW 2026 | litellm 1.107.0 Compromised on PyPI and Docker: Credential Stealer and Hourly Remote Code Loader news 22 min read Python Supply Chain | Library containing malicious payloads, including a credential stealer (`privcollect`) and an hourly remote code execution daemon, was distributed as litellm 1.107.0 via PyPI and Docker Hub. The compromise involved a lookalike GitHub repository and a stolen GitHub organization owner's token. The malware automatically executes on interpreter start, exfiltrating sensitive data like API keys, browser passwords, and SSH keys to `www.litellm.dev`. → stepsecurity.io |
| 2026-11-15 NEW 2026 | Microchip TimePictra hard-coded root password news | A critical vulnerability has been identified in Microchip's TimePictra software, where a hard-coded root password allows unauthorized access. This flaw enables attackers to gain full administrative privileges on affected systems. The password is not user-configurable, meaning all installations are vulnerable unless patched by Microchip. This presents a significant security risk, potentially leading to data breaches or system compromise. Users are advised to monitor for official advisories and updates from Microchip. |
| 2026-11-15 NEW 2026 | Fortinet FortiManager hard-coded credentials news | A security vulnerability has been identified in Fortinet FortiManager, stemming from the use of hard-coded credentials. This flaw allows for potential unauthorized access and control of the affected FortiManager devices. The vulnerability exists in specific versions of FortiManager, and users are advised to update to the latest patched versions to mitigate the risk. This issue highlights the importance of secure coding practices and regular security audits for network management appliances. |
| 2026-11-15 NEW 2026 | OpenBao authentication bypass news AuthN | OpenBao versions prior to 2.6.5 and 2.7.2 have a critical vulnerability allowing an authentication bypass. This flaw could enable unauthorized access to the system. Users are strongly advised to update to a patched version to secure their OpenBao deployments against this exploit. |
| 2026-11-15 NEW 2026 | XWEB EVO hardcoded backdoor password news AuthN | Advisory for GHSA-9vpc-28gc-f73j detailing a hardcoded backdoor password in XWEB EVO versions prior to 5.6.2. This vulnerability, identified as CVE-2026-124004, allows unauthenticated remote attackers to exploit the hardcoded credential "xwebevonewgeneration" to gain full administrative access and control over the application. XWEB EVO reached its end of service in March 2026. |
| 2026-11-15 NEW 2026 | XWEB EVO predictable credential generation news | Advisory for GHSA-hp68-2m45-hqr3 details a predictable credential generation vulnerability in XWEB EVO versions up to 5.8.8. This flaw allows attackers to compute the Linux root password using a fixed prefix and a daily value derived from the MAC address. Versions prior to 5.7.0 had identical root credentials across all units, while later versions still allow derivation. This vulnerability, identified as CVE-2026-124006, enables unauthorized root access to the operating system. |
| 2026-11-15 NEW 2026 | Attacker Used 4,837 Hijacked GitHub Accounts to Run a CI Secret Stealer in 1,806+ Repositories news 6 min read Supply Chain | Library for detecting and mitigating CI secret stealers, similar to those used in the GhostAction campaign. It analyzes attack patterns, including the "docs: dependency audit notes" pull request strategy and the use of npm postinstall hooks, to identify compromised repositories and stealers that exfiltrate GitHub Actions secrets, cloud credentials (AWS, Azure, GCP), and AI service keys. The library also helps detect backdoors created via Entra ID federated credentials and AWS AssumeRole/CreateAccessKey sessions. → thehackernews.com |
| 2026-11-15 NEW 2026 | ARTEX: Attackers Adopted This AI Pentesting Tool in Three Months. Leaked Secrets Are Its Fastest Way In. news 12 min read AI Supply Chain | Library for AI-powered penetration testing, ARTEX, facilitates automated attack planning and execution by AI agents. Attackers rapidly adopted ARTEX, leveraging leaked secrets on platforms like GitHub and guessable development credentials to gain access to sensitive data, including Azure Key Vault contents. Exploitation often involves an AI model independently discovering and utilizing credentials within code repositories and Git history. Backdoored versions of ARTEX have also been distributed, further enabling malicious activities. → stepsecurity.io |
| 2026-11-15 NEW 2026 | Tally Ira’s :- Hardcoded Production Secrets intermediate Bug Bounty | The author discovered hardcoded production API keys within the Tally Ira Android application (com.tally.heytally.apk). This vulnerability allowed for the immediate extraction of all production API keys. The implications of this single finding are significant and cascade into other security concerns within the application. |
| 2026-11-15 NEW 2026 | Information Disclosure — Intercom Identity Verification secret (HMAC key) is hard-coded in… intermediate Bug Bounty | A critical information disclosure vulnerability was found in Intercom's client-side JavaScript. The secret HMAC key, used for identity verification, was hard-coded. This means the key, essential for securing identity verification processes, was exposed directly within the publicly accessible JavaScript code. |
| 2026-11-14 NEW 2026 | The Green Agent POV: Formulating the Fix for an Exposed Container with Sensitive Data Access intermediate 7 min read AuthZ | Library for automated security issue investigation, Green Agent identifies and formulates fixes for exposed containers with sensitive data access. It analyzes attack paths, correlating risks like CVE-2026-111497 ("Broccoli2Shell") with external exposure, insecure Kubernetes configurations, and overprivileged IAM roles. The library then validates exploitability, assesses cloud identity and data risks, and traces runtime assets back to their source code and owners, enabling rapid, zero-downtime mitigations and permanent code fixes. → wiz.io |
| 2026-11-14 NEW 2026 | wfx-loot: A CI Secret Stealer Pushed With Stolen GitHub Credentials news 10 min read Supply Chain | Tool for stealing CI secrets from GitHub repositories, wfx-loot, employed multiple delivery forms including npm hooks to bypass GitHub's workflow execution holds. This technique allowed the stealer to run on developers' machines or within CI pipelines via legitimate package installations, targeting secrets for AWS, Azure, GCP, Docker, Kubernetes, and various package registries. → blog.gitguardian.com |
| 2026-11-14 NEW 2026 | WFX Phantom: The GitHub Actions Secret Stealer That Deletes Its Tracks news 14 min read Supply Chain | Tool for stealing GitHub Actions secrets, named WFX Phantom, operates by pushing a shell script into repositories. This script harvests repository secrets, runner memory, and cloud credentials. Attackers aim to conceal their actions by reverting commits and deleting workflow runs within minutes. Recent versions also exfiltrate data to Telegram and hide the payload within existing shell scripts, complicating detection beyond blocking IP addresses. The tool utilizes valid GitHub tokens, including those issued through the Fly.io GitHub App, to perform these actions. → stepsecurity.io |
| 2026-11-14 NEW 2026 | The key you shipped to every visitor beginner API Sec | A security researcher discovered a live API key with write access embedded directly within a website's JavaScript bundle. This key was accessible to every visitor who downloaded the code. The vulnerability meant that anyone could potentially misuse the API key to alter data or perform other unauthorized actions on the site. The researcher found this by searching the JavaScript code for specific security-related patterns. |
| 2026-11-14 NEW 2026 | Android Pawned — Day 2: Insecure Logging beginner Mobile | Day 2 of "Android Pawned" highlights an "Insecure Logging" vulnerability. A single command and grep filter can expose sensitive information in plaintext within the terminal. This issue arises from developers often prioritizing data security at rest, overlooking the risks associated with logging sensitive data insecurely. |
| 2026-11-14 NEW 2026 | Arize Phoenix: format string injection in applyChatTemplate GraphQL query leaks secrets / JWT signing key intermediate 2 min read AI GraphQL JWT Python | Library vulnerability GHSA-gfv3-m769-5xm3 in Arize Phoenix's `applyChatTemplate` GraphQL endpoint allows format string injection. Exploiting this can disclose server environment variables, cryptographic secrets like the JWT signing key, and internal module attributes. Affected versions range from v13.0.0 to v14.8.0. Recommended mitigations include restricting attribute traversal, adding authentication, and limiting F_STRING format usage. |
| 2026-11-13 2026 | JWT signing key in a JavaScript bundle: case study intermediate 1 min read JWT | Case study detailing an authentication bypass discovered by extracting a hardcoded JWT signing key from a JavaScript bundle. The fallback key, intended as a default, was the actual production HS256 secret, allowing an unauthenticated caller to forge valid session cookies. This enabled minting cookies for any user, granting administrative privileges and the ability to modify user-specific theme settings via a theme API, demonstrating a critical vulnerability missed by standard secret scanners. |
| 2026-11-13 2026 | XWEB EVO hardcoded credential (critical) news | Vulnerability writeup detailing a hardcoded credential flaw in XWEB EVO versions 5.6.2 through 5.8.8. The Administrator account uses a daily password synthesis without a stored secret, allowing attackers to authenticate as Administrator and gain a full admin session if they can derive the MAC-based password. This product line reached its end of service in March 2026. |
| 2026-11-13 2026 | Hijacked GitHub accounts push wfx CI secret stealer to thousands of repos news 13 min read Supply Chain | Library for detecting the "wfx" CI secret stealer, which leverages compromised GitHub accounts to inject malicious code into thousands of repositories. This attack targets sensitive secrets within GitHub Actions, including cloud credentials, package registry tokens, and AI API keys, with the attacker attempting to cover their tracks by reverting changes and deleting logs. The library aims to help organizations identify and mitigate this ongoing threat, which has impacted popular projects like nextcloud/docker and webdriverio. → endorlabs.com |
| 2026-11-13 2026 | wfx-loot: thousands of GitHub accounts hijacked with CI secret-stealing malware news 5 min read Supply Chain | Writeup of wfx-loot, a campaign hijacking thousands of GitHub accounts to inject self-erasing CI secret-stealing malware. Attackers use GitHub Actions and npm install hooks to exfiltrate cloud, package registry, and AI credentials, then revert commits and delete logs. The malware, a single shell script, steals GitHub tokens, AWS, Azure, GCP, npm, PyPI, Packagist, crates.io, Kubernetes, and Vault secrets, with exfiltration shifting from HTTP to Telegram bots. Researchers identified over 66,000 probe pull requests and malicious commits in thousands of public repositories, emphasizing the need for rapid detection and secret rotation. → ox.security |
| 2026-11-13 2026 | Hardcoded credentials in mobile apps: every key is public beginner 8 min read Mobile | Library for finding hardcoded credentials in mobile applications, similar to how researcher Robert Wiggins used AI to discover vulnerabilities in Wyndham and Applebee's apps. This library analyzes decompiled app code to extract keys, passwords, and tokens, including those within third-party SDKs. It then tests these credentials against backend APIs to identify exploitable secrets, addressing the OWASP Mobile Top 10 risk of improper credential usage. |
| 2026-11-13 2026 | HashiCorp Vault integrated storage SSRF news SSRF | HashiCorp Vault versions prior to 1.14.5, 1.13.10, and 1.12.14, as well as Vault Enterprise versions prior to 1.14.5, 1.13.10, and 1.12.14, contain a Server-Side Request Forgery (SSRF) vulnerability in the integrated storage functionality. This vulnerability allows an attacker to trigger requests from the Vault server to arbitrary internal or external endpoints. Users are advised to upgrade to patched versions to mitigate this security risk. |
| 2026-11-12 2026 | AntiFox Targeting AWS Credentials: The Latest Wave of AI Agent Attacks news 10 min read AI | Library for detecting and mitigating AI agent attacks, specifically focusing on the AntiFox framework which targets AWS credentials by chaining credential abuse with reconnaissance. The library aids in discovering exposed secrets, assessing associated permissions, and understanding the broader attack surface presented by agentic offensive security tools like ARTEX and AntiFox, which lower the barrier to entry for credential theft and unauthorized cloud access. → blog.gitguardian.com |
| 2026-11-12 2026 | Stolen AI API Keys: What Railway’s Breach Report Teaches About Agent Credentials news AI API Sec | Railway's recent security incident highlights critical vulnerabilities in managing AI API keys. The company discovered unauthorized access to their systems, which led to the compromise of sensitive AI API keys. This breach underscores the importance of robust security practices for handling agent credentials, especially those granting access to powerful AI models. Organizations must prioritize secure storage, rotation, and monitoring of these keys to prevent similar incidents and protect their AI infrastructure and data from exploitation. The report offers valuable lessons for enhancing security protocols around AI access. |
| 2026-11-12 2026 | AntiFox Targeting AWS Credentials: The Latest Wave of AI Agent Attacks news 10 min read AI | Library for detecting and mitigating AI agent attacks like AntiFox, which targets AWS credentials. This agentic framework, a fork of ARTEX, can discover exposed secrets and leverage them for unauthorized cloud access, often without requiring dedicated secrets scanners. It demonstrates how attackers are lowering the technical barrier to exploiting exposed credentials, necessitating robust Detect, Remediate, and Prevent strategies for credential layer security. → blog.gitguardian.com |
| 2026-11-12 2026 | A Clean Refusal Can Still Leak the Deploy Token news AI | A recent security finding reveals that even when a system correctly rejects a malicious request, it can inadvertently leak a sensitive "deploy token." This occurs because the error response, while functionally correct in refusing access, still includes the deploy token in its output. This means attackers can exploit this behavior to obtain the token, which could grant them unauthorized access to deployment infrastructure. The vulnerability lies in how error messages are constructed, not in the access control logic itself. |
| 2026-11-12 2026 | NextScan 1.1.0 through 1.2.0 hardcoded HTTP Basic Auth credentials news | NextScan versions 1.1.0 through 1.2.0 have a critical security vulnerability due to hardcoded HTTP Basic Authentication credentials. This means that attackers can potentially gain unauthorized access to systems running these versions of NextScan by exploiting these embedded credentials. Users are strongly advised to update to a patched version as soon as possible to mitigate this risk and protect their systems from potential compromise. |
| 2026-11-12 2026 | How an SSRF in a Workflow Tool Led to AWS Credentials and S3 Data Access intermediate SSRF | A Server-Side Request Forgery (SSRF) vulnerability in a workflow tool allowed an attacker to access AWS credentials and sensitive data stored in S3 buckets. The SSRF exploit enabled the attacker to make unauthorized requests from the workflow tool's server, bypassing security controls and ultimately compromising cloud resources. This highlights the critical security implications of SSRF vulnerabilities, especially when they can lead to the exposure of highly sensitive information like cloud credentials and data. |
Frequently Asked Questions
- How do secrets leak into code repositories?
- Secrets commonly leak through developer mistakes: hardcoding API keys during development, committing .env files, leaving credentials in test fixtures, pasting tokens into comments, or including secrets in Docker build arguments. Even if removed in later commits, secrets persist in git history unless the repository is rewritten with tools like git-filter-repo or BFG Repo Cleaner.
- What tools detect leaked secrets?
- TruffleHog and GitLeaks scan git repositories for high-entropy strings and known credential patterns. GitHub Secret Scanning alerts on known token formats from partner services. Pre-commit hooks using detect-secrets or gitleaks can prevent commits containing secrets. For CI/CD, tools like talisman and SpectralOps provide pipeline-level scanning.
- What should you do when a secret is leaked?
- Immediately rotate the compromised credential — assume it has been captured. Revoke the old key, generate a new one, and update all systems using it. Then remove the secret from git history if it was committed. Review access logs for the compromised credential to assess if it was exploited. Finally, implement prevention measures to stop future leaks.
Weekly AppSec Digest
Get new resources delivered every Monday.