gitclassic
Verdict safe, but 2 findings (2 medium), mainly prompt injection, social engineering. Most recent scan 2026-02-06.
This skill presents MODERATE security concerns primarily related to transitive trust and external content handling. The core functionality directs the agent to fetch and parse content from a third-party service (gitclassic.com) without validation mechanisms, creating indirect prompt injection risks if that service is compromised. The skill makes unverifiable performance claims and promotes a commercial service without adequate disclosure. However, the skill is read-only, contains no executable scripts, no data exfiltration code, and no direct prompt injection attempts. The missing 'public.py' reference appears to be a documentation error rather than a security issue. With proper warnings about external content trust and validation mechanisms, this could be a useful utility skill.
Scanner's overall assessment (LLM analyzer)
Second scanner: NVIDIA SkillSpector
SkillSpector found no issues but could not fully inspect the package, usually because SKILL.md refers to a file that is not present, so it fails closed to CAUTION instead of SAFE. Read it as "not fully checked", not as a finding. How the two scanners compare across the whole corpus: Two scanners, one corpus.
What this skill does
Fast, no-JavaScript GitHub browser optimized for AI agents. Browse public repos, read files, view READMEs with sub-500ms load times. PRO adds private repo access via GitHub OAuth.
- Package
- 1 file, 3.1 KB
- SKILL.md
- 404 words, 12+ sections
- Code
- None. Instructions only.
- License
- MIT
- Version
- 1.0.0
- Homepage
- https://gitclassic.com
SKILL.md covers: GitClassic - Fast GitHub Browser for AI Agents · Overview · When to Use · URL Patterns · Repository root · File browser · File contents · User/org profile · Search · Examples · View a repository · Read a specific file
Findings
MEDIUM Transitive Trust Abuse - Following External Web Content Prompt Injectionllm
The skill instructs the agent to fetch and parse content from gitclassic.com, a third-party service, without any validation or security warnings. The agent is directed to trust and follow HTML content from this external source. If gitclassic.com is compromised or serves malicious content, the agent could be subjected to indirect prompt injection attacks embedded in the HTML responses. The skill provides no mechanism to validate the integrity or safety of the content retrieved.
SKILL.md
Instructions direct agent to: 'curl https://gitclassic.com/{owner}/{repo}' and parse responses. No validation of external content is specified. Agent is told to trust 'Clean, semantic HTML' from external source.MEDIUM Potentially Misleading Service Claims Social Engineeringllm
The skill makes strong performance and reliability claims about a third-party service (gitclassic.com) that cannot be verified and may mislead users. Claims include '<500ms load times', 'cached responses' for rate limit avoidance, and 'clean, semantic HTML' that is 'easy to parse'. These claims could be inaccurate or change without notice, potentially causing the agent to make incorrect decisions or set false user expectations. The skill also promotes a paid PRO tier without disclosing the commercial relationship.
SKILL.md
Unverifiable claims: 'Page load <500ms', 'Rate limits: Cached responses', 'AI agent friendly: Clean, semantic HTML'. Commercial promotion: 'PRO subscription ($19/year or $49/lifetime)' without disclosure of author relationship to service.What is in the package
| File | Kind | Size |
|---|---|---|
SKILL.md | doc | 3.1 KB |
External hosts referenced
gitclassic.com
Hostnames that appear in the skill's files. Referencing a host is not a finding; compare against the findings above and the skill's stated purpose.
Package contents recorded 2026-10-22. The scan above ran against the version current at scan time; the two can differ if the skill has since been updated.
What we checked
- Static analyzer. Pattern rules over SKILL.md and bundled files: known-bad commands, exfiltration URLs, secrets, obfuscation.
- Behavioral analyzer. Checks what the skill asks the agent to do against what its description claims it does.
- Llm analyzer. A language model reads the skill and reports intent-level issues the rules cannot express.
- Trigger analyzer. Looks for instructions that fire on conditions the user did not ask for.
- Meta analyzer. A second model pass that validates, deduplicates and re-scores the other analyzers' findings.
What we did not check
- Runtime behaviour. The skill was not executed; findings come from reading its files.
- Anything fetched at install or run time, including remote scripts, package dependencies, and URLs the skill tells the agent to open.
- Content the analyzers cannot read: images, binaries, and some non-English text.
- Known scanner bypass classes. See the methodology and limitations page.
Sources
- Repository: https://clawhub.ai/skills/gitclassic
- Raw scan result: gitclassic-scan.json
- Badge for your README:
- Request a re-scan