marp-deck-workflow
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill coordinates a pipeline where user-provided content and data from intermediate files are used to construct prompts for specialized sub-skills.
- Ingestion points: The skill ingests data from direct user consultations and reads from files such as
definition/deck-definition.md,planning/deck-plan.md, andplanning/deck-visual-design.md. - Boundary markers: There are no explicit delimiters or instructions defined in
SKILL.mdto isolate or warn the agent about potentially malicious instructions embedded within the ingested data. - Capability inventory: The skill has the ability to write to the filesystem, invoke other agent skills (
marp-deck-planner,marp-deck-visual-designer,marp-deck-gen), and execute shell commands. - Sanitization: No sanitization, escaping, or validation of external content is performed before it is interpolated into prompts for downstream tasks.
- [COMMAND_EXECUTION]: The skill is instructed to execute the
task pdfcommand in Step 7 to convert the generated slides into a PDF file. This represents a functional requirement but is a relevant capability for the security profile.
Audit Metadata