Security readout for executives and security teams
Plain-English summary
CVE-2026-22708 affects Cursor before version 2.3. In a specific AI agent configuration, commands expected to be blocked or require approval may still alter the shell environment. That can let malicious prompt content influence later trusted commands. The issue is fixed in Cursor 2.3.
Executive priority
Treat this as a high-priority developer tooling issue, especially where Cursor can run automated terminal actions. The main business risk is AI-assisted workflow manipulation leading to unintended command behavior in developer environments. Upgrade to 2.3 should be tracked to completion.
Technical view
When Cursor Agent runs in Auto-Run Mode with Allowlist mode enabled, certain shell built-ins can execute outside the allowlist and without user approval. An attacker using direct or indirect prompt injection can set, modify, or remove environment variables that affect trusted commands. The CVSS 4.0 score is 7.2 high.
Likely exposure
Exposure is limited to Cursor installations before 2.3 where the Agent is used in Auto-Run Mode with Allowlist mode enabled. Organizations using Cursor for AI-assisted development should prioritize identifying developer workstations or environments with that configuration.
Exploitation context
The source describes exploitation through direct or indirect prompt injection that poisons shell environment variables. The CVSS vector lists attack complexity high and exploitation as unproven. No cited source indicates active exploitation, and the CVE is not marked KEV in the provided data.
Researcher notes
The advisory links this to allowlist bypass and environment-variable poisoning, mapped to CWE-15, CWE-269, CWE-74, CWE-77, CWE-78, and CWE-94. Evidence provided does not include exploit code, observed campaigns, or detailed vulnerable built-ins. Validate only in controlled environments and avoid weaponized testing.
Mitigation direction
- Upgrade Cursor to version 2.3 or later.
- Review Cursor Agent Auto-Run and Allowlist settings on developer systems.
- Disable or restrict risky agent automation until upgraded.
- Check the GitHub advisory for any updated vendor guidance.
Validation and detection
- Inventory Cursor versions and flag installations below 2.3.
- Confirm whether Auto-Run Mode and Allowlist mode are enabled.
- Review developer environment logs for unexpected environment-variable changes.
- Verify upgraded systems report Cursor 2.3 or later.
Public sources used
Generated from the cited source records. This long-tail analysis has not been individually reviewed by a named human.
Conservative CVE-to-ATT&CK context
These mappings and lookup hints may be relevant to the vulnerability behavior, CWE, affected product, or exposure path. Glexia-inferred context is not an official MITRE, ATT&CK, CWE, or CVE Program mapping.
ATT&CK lookup starting points
Use these exact CWE pages and searches to review the Glexia ATT&CK library from this CVE's weakness and description context.
CWE-15: Exact CWE lookup
Use the exact CWE identifier as the starting point before reviewing related ATT&CK behavior. Open the exact CWE lookup page first, then review the ATT&CK searches from that MITRE weakness context. This is a Glexia lookup hint, not an official ATT&CK mapping.
Open ATT&CK lookupCWE-269: Authorization and privilege behavior lookup
Authorization weaknesses can support privilege escalation and valid-account review, depending on exploit path. Open the exact CWE lookup page first, then review the ATT&CK searches from that MITRE weakness context. This is a Glexia lookup hint, not an official ATT&CK mapping.
Open ATT&CK lookupCWE-74: Exact CWE lookup
Use the exact CWE identifier as the starting point before reviewing related ATT&CK behavior. Open the exact CWE lookup page first, then review the ATT&CK searches from that MITRE weakness context. This is a Glexia lookup hint, not an official ATT&CK mapping.
Open ATT&CK lookupCWE-77: Command execution behavior lookup
Command injection weaknesses can lead defenders to review execution techniques and command interpreter telemetry. Open the exact CWE lookup page first, then review the ATT&CK searches from that MITRE weakness context. This is a Glexia lookup hint, not an official ATT&CK mapping.
Open ATT&CK lookupCWE-78: Command execution behavior lookup
Command injection weaknesses can lead defenders to review execution techniques and command interpreter telemetry. Open the exact CWE lookup page first, then review the ATT&CK searches from that MITRE weakness context. This is a Glexia lookup hint, not an official ATT&CK mapping.
Open ATT&CK lookupCWE-94: Code execution behavior lookup
Code execution and unsafe deserialization weaknesses often justify reviewing execution behavior and process telemetry. Open the exact CWE lookup page first, then review the ATT&CK searches from that MITRE weakness context. This is a Glexia lookup hint, not an official ATT&CK mapping.
Open ATT&CK lookup- Severity
- High
- CVSS
- 7.2 (4.0)
- Known Exploited
- No
- Published
Vector: CVSS:4.0/AV:N/AC:H/AT:P/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N/E:U
CNA and ADP enrichment extracted from CVE v5
These fields come from the CVE record and ADP containers, not from Glexia's Take. They preserve time-varying source decisions such as CISA SSVC, KEV status, CVSS metrics, and provider references.
CVSS vector scores
1 official scoreWe collect every scored CVSS vector available in the official CNA and ADP containers. When more than one version is present, the table keeps the source vectors side by side instead of collapsing them into the highest score.
CVSS:4.0/AV:N/AC:H/AT:P/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N/E:U——Primary CVE scoreVulnerability scoring details
Base CVSS 4.0 score
7.2HighVector: CVSS:4.0/AV:N/AC:H/AT:P/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N/E:U
Source materials
- CVE List V5 sourceCVE List V5
- https://github.com/cursor/cursor/security/advisories/GHSA-82wg-qcm4-fp2wCVE reference · x_refsource_CONFIRM
Products and packages named in the record
CWE details
CWE links open Glexia weakness intelligence pages with official CWE context, developer remediation guidance, and related CVE mappings.