MCP Memory Keeper is an MCP server for persistent context management in AI coding assistants. Prior to 0.13.0, context_import in src/index.ts passes the caller-controlled filePath directly to fs.readFileSync without restricting the path to an export directory. An MCP client, including an LLM agent induced to call the tool, can use ../ traversal or an absolute path to target any file readable by the server process. A valid JSON file is parsed and imported into the caller's session, allowing its full contents to be retrieved through context_get or context_export, while JSON.parse errors for non-JSON files can return leading file bytes in a SyntaxError message. The two disclosure modes can expose other exported sessions, JSON credentials or service-account files, environment files, and portions of SSH keys or other local files. This issue is fixed in version 0.13.0.
hashi-vault-js versions prior to 0.5.2 expose sensitive information in error messages. The module rethrows raw Axios errors containing request headers and bodies, which may include Vault tokens and secret data. Logging these errors can lead to plaintext exposure of credentials, risking unauthorized Vault access. The issue is fixed in version 0.5.2.
CVE-2026-66306 is a vulnerability in Microsoft Skype for Business Server 2015 CU13 where error messages generated contain sensitive information. This flaw allows an unauthorized attacker to potentially disclose sensitive information over a network. The issue affects specific versions prior to fixed releases and has a medium severity rating with a CVSS score of 6.5.
CVE-2026-69294 is a vulnerability in Microsoft Windows 10 Version 1809 and later versions that allows an authorized local attacker to obtain sensitive information through error messages generated by Microsoft COM. The flaw involves the generation of error messages containing sensitive data, which could lead to information disclosure. This issue has been officially fixed in multiple Windows 10 builds.
CVE-2026-68886 is a use-after-free vulnerability in the Windows Network Connection Broker component of Microsoft Windows 10 Version 1607 and related builds. An authorized local attacker can exploit this flaw to disclose information. The vulnerability has a medium severity rating with a CVSS score of 5.5. Microsoft has released official fixes for all affected versions.
CVE-2026-67383 is a vulnerability in Microsoft SQL Server 2025 (CU8) where error messages generated by the server may contain sensitive information. This flaw allows an authorized attacker with network access to disclose sensitive information through these error messages. The vulnerability affects specific versions prior to 17.0.4085.5 and 17.0.1135.8. An official fix is available from Microsoft.
CVE-2026-69684 is a vulnerability in Microsoft Windows 10 Version 1607 and later versions that allows an authorized local attacker to disclose sensitive information through error messages generated by Windows Error Reporting. The flaw involves the generation of error messages containing sensitive data, which could be leveraged to gain unauthorized information disclosure. Microsoft has released official fixes for affected versions to address this issue.
CVE-2026-69552 is a vulnerability in Microsoft Windows 10 Print Spooler components that allows an authorized attacker to disclose sensitive information through error messages generated over a network. The flaw involves generation of error messages containing sensitive data, classified under CWE-209. It affects multiple Windows 10 versions prior to specific fixed builds. A patch is available from Microsoft to address this issue.
Generation of Error Message Containing Sensitive Information vulnerability in ash-project ash discloses the stored value of a confirmed field to an actor who fails its confirmation check.
Ash.Resource.Validation.Confirm's atomic implementation (atomic/2 in lib/ash/resource/validation/confirm.ex) built the mismatch error with its value set to the field being confirmed. When the actor supplies only the confirmation argument and not the field itself, value resolves through atomic_ref/2 to the field's current stored value, so the mismatch error echoes that stored value back to the actor. Against a confirmation guarding a sensitive attribute, an actor can submit a deliberately wrong confirmation and read the real value from the returned error. The fix reports the actor-supplied confirmation in the error instead of the stored field value.
This issue affects ash: from 2.17.20 before 3.32.2.
Generation of Error Message Containing Sensitive Information vulnerability in ash-project ash_typescript allows an unauthenticated attacker to read internal application data from an HTTP 500 response body.
When a typed-controller route handler returns anything other than a %Plug.Conn{}, dispatch/3 in lib/ash_typescript/typed_controller/request_handler.ex passes the value to unexpected_return/2, which interpolates inspect(value, limit: 50) directly into the response message. The limit option bounds elements per collection rather than the term as a whole, so a handler falling through with a term such as {:error, %User{}} or a changeset serialises its full field set, including hashed passwords, tokens, and tenant identifiers, into the JSON error returned to the caller.
This contradicts the module's own posture elsewhere: the rescue clause gates Exception.message/1 behind AshTypescript.typed_controller_show_raised_errors?/0 and otherwise returns a generic message, while this path is ungated and always echoes.
This issue affects ash_typescript: from 0.15.0 before 0.18.0.
Generation of Error Message Containing Sensitive Information vulnerability in ash-project ash_typescript allows an unauthenticated attacker to receive unredacted internal error data by provoking an error shape the configured error handler does not match.
apply_error_handler/3 in lib/ash_typescript/rpc/errors.ex is the only hook an application has for redacting or suppressing errors before they reach the client, with a nil return dropping the error entirely. Its rescue clause logs a warning and then returns the original, pre-handler error map. Error handlers are conventionally written as pattern-matching functions over expected error shapes, so an unmatched shape raises FunctionClauseError and the raw transformed error, including any secrets carried in vars, is emitted instead. An intent to suppress an error becomes an intent to publish it. The rescue catches exceptions only, so a handler that throws or exits still propagates.
This issue affects ash_typescript: from 0.8.0 before 0.18.0.
CVE-2026-82727 is a vulnerability in ash-project's ash_phoenix component where error messages generated during form processing leak sensitive information. Specifically, when a union sub-form receives an unexpected _union_type value, the entire submitted parameter map, including secrets like passwords, is embedded in the exception message. This sensitive data is exposed in logs, crash reports, and developer error pages because the message bypasses Phoenix's parameter filtering. The issue affects versions from 1.2.17 up to but not including 2.3.25.
CVE-2026-75760 is a vulnerability in ash-project's ash_ai component that causes error messages to leak sensitive information. When the embedding provider call fails in AshAi.Changes.Vectorize, the error message includes unsanitized details such as the request URL, provider response body, and potentially the Authorization header containing the API key. This information is exposed to users via validation errors. The issue affects versions from 0.1.0 up to but not including 1.0.0. The vulnerability has a high severity score of 7.1.
CVE-2026-82580 is a vulnerability in ash-project ash_ai where internal error messages containing sensitive information are disclosed to chat users. This occurs because exceptions raised during tool execution are serialized and sent back to the user without filtering. The issue affects versions from 0.6.0 up to but not including 1.0.0. The vulnerability has a medium severity score of 5.3. No official patch or remediation guidance is currently available.
CVE-2026-78693 is a vulnerability in ash-project's ash_graphql component that causes error messages to expose internal field names despite application-level redaction configurations. This occurs because the error handling mechanism reinjects internal attribute and argument names into error paths when no field name mapping is configured, leading to leakage of sensitive internal information. The issue affects versions from 1.9.0 up to but not including 1.11.0. The vulnerability has a medium severity score of 6.9 and does not currently have an official patch or remediation guidance.
CVE-2026-47893 is a high severity vulnerability in the Spring Framework affecting multiple versions. It involves the generation of error messages in Spring WebFlux applications that support WebSocket connections, where request headers may be included in exception reasons, potentially exposing sensitive user information. The vulnerability affects versions from 5.2.25.RELEASE and earlier up to 7.0.8. There is no explicit vendor advisory or patch information available at this time.
CVE-2026-59271 is a vulnerability in Spring AMQP where the RabbitMQ management aliveness check failure causes the configured admin password to be exposed in cleartext within the exception message. This affects multiple versions of Spring AMQP including 2.4.18 and earlier, 3.2.0 through 3.2.12, 4.0.0 through 4.0.4, and exactly 4.1.0. The vulnerability has a medium severity with a CVSS score of 5.3. No official patch or remediation guidance is currently provided.
HCL BigFix Quantum Risk Analyzer version 2.0.1.47 has a vulnerability where overly descriptive error messages are generated upon receiving malformed input. This can aid attackers in reconnaissance and refining fuzzing techniques. The issue is identified as CWE-209, involving the generation of error messages containing sensitive information. The CVSS score is 3.9, indicating a low severity level.
CVE-2026-8173 is a medium severity vulnerability in the Murrelektronik Xelity 4TX M GE switch web GUI version 2.1.0. It involves improper error message generation that allows an unauthenticated attacker with network access to the web interface to retrieve MAC addresses logged by the device. This occurs when an authenticated administrator uses the 'Copy learned MAC Addresses' function, and the error messages leak sensitive MAC address information.
A vulnerability in Combodo iTop prior to version 3.2.3 allows sensitive information to be disclosed through error messages. This issue is classified under CWE-209, which involves generation of error messages containing sensitive information. The vulnerability has a low severity score and has been fixed in version 3.2.3.