Skills
Navigate
skills/jwynia/the-kepler-testimonies/dialogue/Gen Agent Trust Hub

dialogue

Pass

Audited by Gen Agent Trust Hub on Sep 30, 2026

Risk Level: SAFECOMMAND_EXECUTIONPERSISTENCEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes local Deno scripts (voice-check.ts, dialogue-audit.ts) for analysis. Execution is restricted with the --allow-read flag, limiting access to reading files.\n- [PERSISTENCE]: The skill maintains session state by writing results to .dialogue-output.md and context/output-config.md. This behavior is transparently documented and serves the skill's primary purpose.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted text files through analysis scripts.\n
  • Ingestion points: Deno.readTextFile (scripts/dialogue-audit.ts, scripts/voice-check.ts)\n
  • Boundary markers: None\n
  • Capability inventory: deno run --allow-read, file writes to project-local markdown and config files\n
  • Sanitization: None
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 30, 2026, 03:34 AM