AI Stack Alert High CVE-2026-117558 For AppSec, AI Eng, SOC / IR, CISO AI-assisted
CVE-2026-117558: OS command injection in ai-sdlc plugin MCP server through 0.31.2
The ai-sdlc plugin MCP server through version 0.31.2 has an OS command injection flaw in its repo argument. The unvalidated value is interpolated into gh issue view or gh pr view commands run via execSync(). The reported CVSS base score is 8.5.
Why it matters
An AI agent that is manipulated through prompt injection can run arbitrary commands as the MCP server user. Any organization running this plugin with agent access to MCP tools is exposed.
Do this
- Identify all hosts and developer workstations running the ai-sdlc plugin at version 0.31.2 or earlier.
- Check the vendor or project for a fixed release. The NVD entry does not name one. Upgrade or remove the plugin until a fix is confirmed.
- Limit which agents can call the ai-sdlc MCP server tools, and review the server user's privileges.
- Review logs of gh issue view and gh pr view invocations for repo values containing shell metacharacters or unexpected characters.
Exploit Clock
Exploitation unlikely (for now)
- EPSS 0.69%
Signals
- NoCISA KEV
- 8.5CVSS
- 0.69%EPSS
- NoPublic PoC
- 34/100Priority
- 70%Confidence
Evidence
“The ai-sdlc plugin MCP server through 0.31.2 contains an OS command injection vulnerability that allows attackers to execute commands by supplying shell metacharacters in the repo argument.”
“The unvalidated repo value is interpolated into gh issue view or gh pr view commands run via execSync(), letting prompt-injected agents execute commands as the MCP server user.”
Sources
Affected / related
ai-sdlc plugin, OS command injection
Published 2026-10-27 14:16 UTC · updated 2026-11-08 02:42 UTC · AI-assisted summary, checked against quoted sources · How we score