skills/ilandahan/aid/learning-mode/Gen Agent Trust Hub

learning-mode

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill employs strong directive language and override markers to ensure methodology adherence, explicitly stating that the behavior is 'NOT optional' and 'MUST' be followed.
  • [PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by directing the agent to 'capture learnings' from user feedback and apply them to future interactions. * Ingestion points: User responses to feedback requests as described in SKILL.md and SIMULATION.md. * Boundary markers: There are no instructions or delimiters provided to isolate user input or treat it as untrusted data during the learning process. * Capability inventory: The agent is authorized to modify its future reasoning, prioritization, and technical approach based on user feedback. * Sanitization: No mechanisms for sanitizing or validating user-provided feedback are defined before the agent integrates the feedback into its behavioral profile.
  • [NO_CODE]: The skill contains no executable scripts, binaries, or automated installation steps, consisting entirely of instructional markdown files.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 03:46 PM