Is superdogidog/Learnapp safe?
Yes. superdogidog/Learnapp is safe to install. Oathe's behavioral security audit gave the Learnapp skill by superdogidog a trust score of 92/100 with 5 findings, none critical or high.
https://github.com/superdogidog/Learnapp
Is superdogidog/Learnapp safe to install?
superdogidog/Learnapp is a legitimate React/Vite Chinese language learning web application with 1310 committed pinyin MP3 files, a service worker for offline caching, and a responsive UI. Its SKILL.md is completely empty, meaning it has no agent-facing instruction surface and cannot function as an LLM skill or plugin. No malicious behavior, prompt injection, exfiltration attempts, or canary violations were detected during the audit.
What security issues were found in superdogidog/Learnapp?
Category Scores
Findings (5)
INFO SKILL.md is empty — no skill functionality present -2 ▶
The repository contains no SKILL.md content. This repository is a standalone React web application for learning Chinese (pinyin audio, character recognition, listening exercises), not an LLM agent skill or plugin. There is no instruction surface that could be injected into an agent's system prompt.
Evidence: SKILL.md content: (empty string)
LOW Third-party translation API receives user input -15 ▶
translationHelper.js constructs a URL containing user-supplied Chinese text and POSTs it to https://api.mymemory.translated.net. While this is a legitimate public translation service and the data is user-initiated, it does represent an external data channel. In an agent context where an LLM might call this code, it could inadvertently send agent-visible content to a third party.
Evidence: const url = `https://api.mymemory.translated.net/get?q=${encodedText}&langpair=en|ru`; (translationHelper.js)
LOW downloader.py present — content unverified -10 ▶
A Python script named downloader.py is committed to the repository. Its full content was not captured in the evidence. Based on repository context (1310 pinyin MP3 files committed), it is almost certainly a one-time utility script used to download audio files. It does not run automatically via npm lifecycle scripts or git hooks.
Evidence: File present: /home/oc-exec/skill-under-test/superdogidog/Learnapp/downloader.py — content not shown in audit evidence
INFO Normal clone behavior — only GitHub contacted -5 ▶
Network monitoring showed a single external TCP connection to 140.82.121.4:443 (github.com) during the clone operation. No other external hosts were contacted. No new listening ports were opened. The connection state before and after the install is identical save for a new SSH session from the audit runner.
Evidence: 04:10:18.413435 ens3 Out IP 10.0.2.15.45212 > 140.82.121.4.443: tcp 0 (git clone to github.com)
INFO Canary file accesses predate clone — attributable to audit infrastructure 0 ▶
Filesystem events show accesses to .env, .ssh/id_rsa, .aws/credentials, and other honeypot files at timestamp 04:10:12. The git clone of this repository did not begin until 04:10:18. These accesses are therefore from the audit orchestration infrastructure (establishing canary baselines), not from the skill under test. All canary files are confirmed intact by the monitoring system.
Evidence: 04:10:12 OPEN /home/oc-exec/.env (before clone at 04:10:18); Monitoring: ✅ All canary files intact
Should I install superdogidog/Learnapp?
Oathe's verdict for superdogidog/Learnapp is SAFE with a trust score of 92/100. Recommendation: Install with caution.