CVE-2026-61447: PraisonAI: CodeAgent Executes LLM-Generated Code Without Sandboxing and Leaks All Environment Secrets

Should you act?

Critical and remotely reachable

No known exploitation yet, but it is network-reachable without credentials. This signal is computed from CISA KEV, EPSS and CVSS data; it is not an editorial verdict.

• Not in CISA KEV, but CISA’s SSVC assessment records a public proof-of-concept, so attack code exists.
• EPSS 2.5%, percentile 84: FIRST’s model estimates a 2.5% chance of exploitation activity in the next 30 days, higher than 84% of all scored CVEs (elevated). EPSS is a probability, not a verdict.
• Remotely reachable with no login and no user interaction. If the vulnerable service is exposed to the internet, prioritize it.
• Only applies if your dependency tree includes praisonaiagents (PyPI) <= 1.6.77, fixed in 1.6.78. Check your lockfile, including transitive dependencies.

What it is

PraisonAI before 1.6.78 contains a remote code execution vulnerability in CodeAgent.executepython() that executes LLM-generated Python code without AST validation, import restrictions, or sandbox enforcement. Attackers can influence LLM output through prompt injection to exfiltrate all environment secrets and execute arbitrary code on the host system.

From the GitHub advisory

Summary CodeAgent.executepython() executes LLM-generated Python code in a subprocess with the complete parent-process environment (os.environ.copy()), zero AST validation, zero import restrictions, and no sandbox enforcement — even when CodeConfig(sandbox=True) is explicitly set. This allows an attacker who can influence LLM output (via prompt injection in agent input, tool results, or ingested content) to exfiltrate all environment secrets (API keys, database credentials, cloud tokens) and execute arbitrary code on the host.

Details

src/praisonai-agents/praisonaiagents/agent/codeagent.py (lines 253–308):

Key issues:

1. Environment leak: os.environ.copy() passes every environment variable — OPENAIAPIKEY, DATABASEURL, AWS credentials, etc. to the subprocess. By contrast, the sandboxed executecode tool in pythontools.py uses env={} (empty environment).

Weakness type: CWE-94, CWE-200.

Product: praisonaiagents (PyPI) · Assigned by: VulnCheck · Published: 2026-07-11

Affected and fixed versions

Severity scores

• Attack vector: network (remotely reachable)
• Attack complexity: low
• Attack requirements: none
• Privileges required: none (no login)
• User interaction: none

CVSS measures how bad a flaw is if exploited, not how likely exploitation is. Read it next to the exploitation data below.

Exploitation and urgency

• CISA KEV: not listed (no confirmed in-the-wild exploitation known to CISA).
• CISA SSVC assessment: exploitation poc, automatable yes, technical impact total.
• FIRST EPSS (model of 2026-10-21): 2.49% probability of exploitation activity in the next 30 days; percentile 84.

Timeline

• 2026-07-11 — CVE record published by VulnCheck
• 2026-07-13 — CVE record last updated
• 2026-10-08 — GitHub-reviewed advisory GHSA-2xv2-w8cq-5gxw published (critical)

Sources and references

Original sources first. Version and fix details should always be confirmed against the vendor or project advisory.

Vendor and project advisories, patches

1. GitHub Security Advisory (GHSA-2xv2-w8cq-5gxw) github.com

Vulnerability records

1. GitHub Advisory GHSA-2xv2-w8cq-5gxw github.com
2. CVE.org record CVE-2026-61447 cve.org
3. NVD entry CVE-2026-61447 nvd.nist.gov

Other references from the CVE record and advisory

1. VulnCheck Advisory: PraisonAI before 1.6.78 Remote Code Execution via CodeAgent vulncheck.com

Critical PyPI advisories · Today’s Security Digest