NYC
skills/letta-ai/skills/raman-fitting/Gen Agent Trust Hub

raman-fitting

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • PROMPT_INJECTION (SAFE): No instructions attempting to override agent behavior or bypass safety filters were detected. The content is strictly technical and educational.
  • DATA_EXPOSURE & EXFILTRATION (SAFE): No hardcoded credentials, sensitive file paths, or unauthorized network operations were identified. The data parsing instructions refer to local spectroscopic data provided by the user.
  • OBFUSCATION (SAFE): No hidden characters, encoded commands, or deceptive text techniques (Base64, zero-width characters, homoglyphs) were found.
  • EXTERNAL_DOWNLOADS (SAFE): No remote script downloads or unauthorized package installations were detected. It references the well-known scientific library 'scipy'.
  • COMMAND_EXECUTION (SAFE): The provided Python snippets are for data processing and visualization; they do not invoke shell commands, sub-processes, or system-level modifications.
  • INDIRECT_PROMPT_INJECTION (SAFE): While the skill describes data ingestion (parsing raw text files), it lacks dangerous capabilities like network access or file-writing that could be exploited via malicious data payloads. It explicitly recommends manual data verification and physical sanity checks.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:04 PM