qa-test-planner
Audited by Socket on Sep 15, 2026
2 alerts found:
Securityx2No indicators of classic package-supplied malware (no network activity, credential theft, persistence, or system modification beyond generating an output file). However, the script contains two significant security weaknesses: (1) use of `eval` with user-controlled input, which can enable arbitrary command execution under the right conditions; and (2) unconstrained `OUTPUT_DIR` from the command-line argument, enabling write/overwrite of arbitrary files within the executing user's permissions. Even though it primarily generates Markdown, unescaped user content can also create downstream injection risk depending on how the generated files are consumed.
This is a local interactive bug-report generator with a serious security weakness: it uses `eval` on user-provided input in `prompt_input`, which can allow arbitrary shell command execution if an attacker can influence the entered text (or if the script is run in a context where input is not fully trusted). Additionally, the output directory can be controlled via the first CLI argument, enabling potentially unsafe file writes in attacker-chosen locations (subject to permissions). There are no clear signs of intentional malware (no exfiltration/persistence/network behavior), but the `eval` pattern makes the script risky to run with untrusted input.