character-arc
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input (character names) and uses it to construct file paths for persisting analysis, which could lead to path traversal if not handled carefully by the agent.\n
- Ingestion points: Character names and story details provided by the user in the interactive conversation as described in the Output Discovery and File Naming sections.\n
- Boundary markers: Absent; the skill does not define specific delimiters or instructions for the agent to separate user input from system instructions or ignore instructions embedded in user content.\n
- Capability inventory: File system write access is required to save the character arc analysis to files such as {character-name}-arc-{date}.md.\n
- Sanitization: Absent; no explicit instructions are provided in the skill for the agent to validate, sanitize, or escape the character name before using it in the file naming pattern.
Audit Metadata