create-html-deck
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run a bundled Python script,
scripts/check_deck.py, to perform structural validation on the generated HTML presentations. This is an intended functional part of the skill. - [DATA_EXFILTRATION]: The
check_deck.pyscript resolves and verifies the existence of local file paths referenced in HTML<img>tags. This behavior could be used to probe for the existence of sensitive files if the agent is directed to process a malicious HTML file, though it does not exfiltrate the contents of those files. - [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it is designed to ingest and process existing HTML presentations.
- Ingestion points: Instructions in
SKILL.mddirect the agent to inspect the HTML, CSS, and JavaScript of existing files before performing edits. - Boundary markers: The skill lacks explicit boundary markers or instructions to ignore embedded commands within the processed HTML files.
- Capability inventory: The agent uses the
scripts/check_deck.pyscript for analysis alongside standard file-system and terminal capabilities. - Sanitization: No validation or sanitization is performed on the untrusted HTML content prior to ingestion into the agent's context.
Audit Metadata