first-cut-davinci-dialogue
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data in the form of word-timed transcripts and audio waveforms to automate editing decisions.
- Ingestion points: Found in SKILL.md and references/waveform-word-timing.md, where the skill reads transcript and waveform data.
- Boundary markers: The instructions lack explicit delimiters or warnings to ignore instructions that might be embedded within the transcript text.
- Capability inventory: The skill utilizes the Resolve MCP tool and others to execute destructive timeline edits (cuts, ripple deletes) and apply transitions.
- Sanitization: No sanitization or validation logic is defined for the textual content of the transcripts.
Audit Metadata