internal-linker
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFENO_CODEPROMPT_INJECTION
Full Analysis
- [NO_CODE]: The skill is comprised entirely of Markdown-based instructions and logic descriptions. It does not contain any executable scripts, binaries, or package dependencies that could facilitate runtime attacks.
- [SAFE]: Analysis of the auditing workflow shows it is limited to processing structural information from local project files (such as H1 tags and internal URLs). There are no patterns suggesting data exfiltration, credential harvesting, or unauthorized privilege escalation.
- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it instructs the agent to ingest and process data from potentially untrusted project files (Category 8). However, the risk is negligible as the skill lacks write or network-based capabilities.
- Ingestion points: Local project files such as page.tsx and post.md.
- Boundary markers: None specified in the instructions.
- Capability inventory: Limited to metadata extraction, link mapping, and report generation; no file-writing or network-sending tools are used.
- Sanitization: No explicit sanitization or escaping of the ingested content is described.
Audit Metadata