skillpper-saver
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from repository files and browser automation (DOM text, headings, accessibility tree) while maintaining capabilities to execute shell commands and write local files. The instructions do not specify the use of boundary markers or delimiters for this external data, nor do they include requirements for sanitization or escaping before interpolation into the agent's context.
- Ingestion points:
SKILL.md,references/automation-and-tools.md(repository files, web page content, and error logs). - Boundary markers: Absent; the skill does not instruct the agent to use delimiters or ignore-embedded-instruction warnings when processing external data.
- Capability inventory: Shell command execution (
grep,rg,find,glob,gh), browser automation (Playwright, Puppeteer), and file system writes (scratch/error.log, handoff notes). - Sanitization: Absent; no mention of filtering, escaping, or validating external content before it enters the session context.
Audit Metadata