skills/djalmajr/skills/wiki-lint/Gen Agent Trust Hub

wiki-lint

Pass

Audited by Gen Agent Trust Hub on May 12, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: No malicious patterns or unauthorized exfiltration attempts were detected. External references target reputable sources.
  • [EXTERNAL_DOWNLOADS]: Mentions the QMD tool from github.com/tobi/qmd for semantic searching, which is an established project for these tasks.
  • [DATA_EXFILTRATION]: Accesses local directories (wiki/, raw/) for analysis but lacks network access for external data transmission.
  • [PROMPT_INJECTION]: As the skill processes markdown content, it maintains an indirect prompt injection surface. (1) Ingestion points: Markdown files within wiki/ and raw/ folders (SKILL.md). (2) Boundary markers: None defined to isolate data from instructions. (3) Capability inventory: Can perform search queries and write updates to index.md, log.md, and wiki page metadata (SKILL.md). (4) Sanitization: No sanitization logic for ingested content was observed. Note: The instruction to summarize automatic fixes rather than listing them in detail is a standard behavior for linting tools.
Audit Metadata
Risk Level
SAFE
Analyzed
May 12, 2026, 09:19 PM