CVE-2026-5972

CVE-2026-5972 affects FoundationAgents MetaGPT (up to 0.8.1). The vulnerability lies in Terminal.run_command within metagpt/tools/libs/terminal.py, where input handling allows os command injection. This enables remote exploitation as described in multiple sources. Patch identifier d04ffc8dc67903e...

CVE-2026-5972

A vulnerability has been found in FoundationAgents MetaGPT up to 0.8.1. This issue affects the function Terminal.runcommand in the library metagpt/tools/libs/terminal.py. The manipulation leads to os command injection. Remote exploitation of the attack is possible. The exploit has been disclosed ...

CVE-2026-5972 FoundationAgents MetaGPT terminal.py Terminal.run_command os command injection

A vulnerability has been found in FoundationAgents MetaGPT up to 0.8.1. This issue affects the function Terminal.runcommand in the library metagpt/tools/libs/terminal.py. The manipulation leads to os command injection. Remote exploitation of the attack is possible. The exploit has been disclosed ...

UBUNTU-CVE-2026-5441

An out-of-bounds read vulnerability exists in the DecodePsmctRle1 function of DicomImageDecoder.cpp. The PMSCTRLE1 decompression routine, which decodes the proprietary Philips Compression format, does not properly validate escape markers placed near the end of the compressed data stream. A crafte...

PYSEC-2026-21

Apache Airflow versions 3.0.0 through 3.1.8 DagRun wait endpoint returns XCom result values even to users who only have DAG Run read permissions, such as the Viewer role.This behavior conflicts with the FAB RBAC model, which treats XCom as a separate protected resource, and with the security mode...

Missing Authorization

Overview Affected versions of this package are vulnerable to Missing Authorization through the waitdagrununtilfinished handler in airflow-core/src/airflow/apifastapi/coreapi/routes/public/dagrun.py. An attacker can read task result values by sending a GET request to the DAG run wait endpoint with...

PraisonAI 参数注入漏洞

PraisonAI is a low-code multi-agent collaboration framework developed by Mervin Praison. Versions of PraisonAI prior to 4.5.128 contained a parameter injection vulnerability. This vulnerability stemmed from the deploy.py script, which did not validate the values containing commas when constructin...

PraisonAI 操作系统命令注入漏洞

PraisonAI is a low-code multi-agent collaboration framework developed by Mervin Praison. Versions of PraisonAI prior to 1.5.128 contained a vulnerability related to operating system command injection. This vulnerability stemmed from the memory hook executor directly passing user-controlled comman...

MetaGPT 操作系统命令注入漏洞

MetaGPT is a multi-agent framework developed by MetaGPT Inc. Versions of MetaGPT 0.8.1 and earlier contained a vulnerability related to operating system command injection. This vulnerability stemmed from the Bash.run function in the metagpt/tools/libs/terminal.py library, which could allow attack...

PT-2026-31707

Name of the Vulnerable Software and Affected Versions FoundationAgents MetaGPT versions up to 0.8.1 Description A flaw exists in the Terminal.run command function within the metagpt/tools/libs/terminal.py library. This allows for os command injection, potentially enabling remote exploitation. The...

CVE-2026-40113: Improper Neutralization of Argument Delimiters in a Command

PraisonAI is a multi-agent teams system. Prior to 4.5.128, deploy.py constructs a single comma-delimited string for the gcloud run deploy --set-env-vars argument by directly interpolating openaimodel, openaikey, and openaibase without validating that these values do not contain commas. gcloud use...

PT-2026-31630

An out-of-bounds read vulnerability exists in the DecodePsmctRle1 function of DicomImageDecoder.cpp. The PMSCT RLE1 decompression routine, which decodes the proprietary Philips Compression format, does not properly validate escape markers placed near the end of the compressed data stream. A craft...

MetaGPT 操作系统命令注入漏洞

MetaGPT is a multi-agent framework developed by MetaGPT Inc. Versions of MetaGPT 0.8.1 and earlier contained a vulnerability related to operating system command injection. This vulnerability stemmed from the Terminal.runcommand function in the metagpt/tools/libs/terminal.py library, which could...

PT-2026-31782

Name of the Vulnerable Software and Affected Versions PraisonAI versions prior to 4.5.128 Description PraisonAI is a multi-agent teams system. The deploy.py script constructs a comma-delimited string for the gcloud run deploy --set-env-vars argument by directly interpolating openai model, openai...

Information Exposure

Overview Affected versions of this package are vulnerable to Information Exposure via the runPlaygroundServer process in cmd/run/run.go and the playground configuration in pkg/server/config/config.go. An attacker can recover the preshared API key by sending an unauthenticated request to the...

Inefficient Algorithmic Complexity

Overview @chenglou/pretext is a Fast, accurate & comprehensive text measurement & layout Affected versions of this package are vulnerable to Inefficient Algorithmic Complexity through the isRepeatedSingleCharRun function during text analysis. An attacker can cause significant performance...

GHSA-5478-66C3-RHXR Pretext: Algorithmic Complexity (DoS) in the text analysis phase

isRepeatedSingleCharRun in src/analysis.ts line 285 re-scans the entire accumulated segment on every merge iteration during text analysis, producing On² total work for input consisting of repeated identical punctuation characters. An attacker who controls text passed to prepare can block the main...

Emissary has GitHub Actions Shell Injection via Workflow Inputs

Summary Three GitHub Actions workflow files contained 10 shell injection points where user-controlled workflowdispatch inputs were interpolated directly into shell commands via $ expression syntax. An attacker with repository write access could inject arbitrary shell commands, leading to reposito...

OpenClaw Authorization Problem Vulnerability (CNVD-2026-16621)

OpenClaw is a command line tool for rights management. A security vulnerability exists in versions of OpenClaw prior to 2026.3.11, which stems from the failure of the system.run approval function to properly bind variable file operands for specific script runners such as tsx, jiti, and others. An...

uac 操作系统命令注入漏洞

UAC is a Unix system forensics and incident response tool developed by Thiago Canozzo Lahr. Versions of UAC prior to 3.3.0-rc1 contained a vulnerability related to operating system command injection. This vulnerability stemmed from the runcommand function, which directly passed the constructed...