term-decision
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local Python scripts using the
uvtool to manage and verify glossary entries. - Evidence: Step 4 and Step 5 in
SKILL.mddemonstrate shell commandsuv run python scripts/term_edit.pyanduv run python scripts/term_read.pyused to persist decisions and verify consistency. - [INDIRECT_PROMPT_INJECTION]: The skill processes external, potentially untrusted markdown content, creating a surface for indirect prompt injection attacks.
- Ingestion points: As described in
SKILL.md, the skill collects terminology candidates from source glossary pages, terminology pages, and index/appendix lists within extracted markdown files. - Boundary markers: Absent. The instructions do not specify any delimiters or warnings to the agent to disregard instructions embedded within the source markdown files.
- Capability inventory: The skill can execute local Python scripts (
term_edit.py,term_read.py) that read from and write to project configuration files (glossary.json,style-decisions.json). - Sanitization: Absent. There is no mention of sanitizing or escaping the terminology extracted from source documents before it is used as arguments in shell commands.
Audit Metadata