/

CVE-2026-78136

Severity:

HIGH RISK

(7.8)

Eval Injection (CWE-95) in CHIRP Kenwood ITM Driver

CVE-2026-78136

Deferred

Vulnerability type

Category:

Code Injection

CWE:

CWE-95

Impact:

Arbitrary Code Execution

Severity:

HIGH RISK

(7.8)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

A code injection vulnerability in CHIRP allows crafted CSV data to reach an unsafe eval operation in the Kenwood ITM driver's _clean_tmode function. If an attacker can supply or induce processing of malicious CSV content, the application may execute unintended Python expressions with the privileges of the CHIRP process.

Affected context

The issue occurs in CHIRP's Kenwood ITM driver when imported or otherwise processed CSV data is passed through _clean_tmode. Deployments are at risk when CSV files can be obtained from untrusted sources, uploaded by users, synchronized from external systems, or processed without adequate content validation.

Exploit preconditions

  • Attack surface: CSV import or processing through the Kenwood ITM driver
  • Required access: Ability to provide, modify, or cause processing of a crafted CSV file
  • Trigger point: Handling of the affected data in _clean_tmode
  • Why exploitation is possible: CSV-derived content reaches an evaluation mechanism that interprets input as executable Python expression content rather than treating it solely as data

Why this happens

The underlying bug is dynamic evaluation of input-derived text. Use of eval on values originating from CSV data allows attacker-controlled content to be interpreted as code, matching CWE-95: Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection'). The insecure pattern is treating external data as executable logic instead of parsing it with explicit, non-executable conversion rules.

Potential impact

  • An attacker may execute arbitrary code in the CHIRP process context
  • Local files or configuration data could be read, altered, or deleted
  • Sensitive credentials available to the process could be exposed
  • The host environment could be modified or used for further compromise
  • Imported radio configuration data could be corrupted or manipulated

General mitigation guidance

Mitigation generally focuses on removing dynamic evaluation from data-processing paths.

Common defensive patterns include:

  • Replacing eval with explicit parsing, mapping, or type-conversion logic
  • Treating CSV fields as untrusted data regardless of their source
  • Applying allowlisted formats and strict schema validation before transformation
  • Rejecting unexpected characters, structures, and values without attempting interpretation
  • Processing imported files with least-privileged operating-system accounts
  • Isolating file-processing workflows and limiting access to sensitive resources
  • Keeping imported data and executable application logic in separate trust domains

How teams detect issues like this

Detection can include static analysis rules that identify eval or equivalent dynamic execution APIs operating on file, CSV, or user-controlled input. Code review should examine data flows from import functions into expression evaluators, conversion routines, and driver-specific cleanup methods.

Policy enforcement may require approved parsing libraries, prohibit dynamic evaluation in ingestion code, and add regression tests for hostile and malformed CSV values. Automated code review and security scanning tools can help identify these patterns during development.

Sources

SourceLink
github.comgithub.com/cduram/CHIRP-CodeExecution_via_Malicious_ImageFile
github.comgithub.com/kk7ds/chirp/commit/39178dbfc4fece083ab9ed20286d6ae3a91a718e

Frequently asked

questions

Exploitability depends on how CSV data reaches CHIRP. It may be exploitable through a remote workflow if an exposed service or integration accepts attacker-controlled files, but a local file-processing workflow generally requires local or supply-chain access.

An attacker needs the ability to provide, alter, or cause processing of crafted CSV data through the affected Kenwood ITM driver.

Remove dynamic evaluation from the CSV handling path and replace it with explicit, allowlisted parsing and conversion logic.

Treat imported files as untrusted, enforce strict schemas, prohibit evaluation of external data, review input-to-execution data flows, and run file processing with minimal privileges.

Find out what's already
exploitable in your codebase.