accelint-skill-manager

Pass

Audited by Gen Agent Trust Hub on Mar 19, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [SAFE]: Analysis of the skill files confirms they contain documentation, templates, and architectural guidance for AI agents. No executable malicious code, obfuscation, or persistence mechanisms were detected.
  • [EXTERNAL_DOWNLOADS]: The README provides installation instructions using npx from the author's official GitHub repository (github.com/gohypergiant/agent-skills). This is documented neutrally as it targets a verified vendor source and follows standard framework delivery patterns.
  • [PROMPT_INJECTION]: The skill includes a workflow for auditing existing skills, creating an indirect prompt injection surface when processing untrusted content. 1. Ingestion points: Reads local skill files from .claude/skills and ~/.claude/skills during the initialization and audit steps in SKILL.md. 2. Boundary markers: The skill does not explicitly instruct agents to use delimiters to isolate audited content from instructions. 3. Capability inventory: Instructions in SKILL.md and references/file-system.md provide guidance for directory creation and file system modifications. 4. Sanitization: No explicit filtering or sanitization of input skill content is mentioned. This risk factor is documented as a inherent property of the tool's primary purpose for skill management.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 19, 2026, 07:52 PM