skills/garrytan/gbrain/briefing/Gen Agent Trust Hub

briefing

Pass

Audited by Gen Agent Trust Hub on Apr 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions do not contain any malicious patterns such as prompt injection, persistence mechanisms, or obfuscation. Its primary purpose is to organize internal data for user review.
  • [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes data from external sources like brain pages and calendar invites, though its impact is mitigated by a read-only toolset.
  • Ingestion points: Data enters the context via gbrain search, gbrain get, gbrain query, and gbrain get_timeline tools as described in the 'GBrain-Native Context Loading' section of SKILL.md.
  • Boundary markers: Absent. There are no explicit instructions to use XML tags or delimiters to isolate data retrieved from the brain from the agent's instructions.
  • Capability inventory: The skill is restricted to information retrieval tools (search, query, get_page, list_pages, get_timeline). It lacks capabilities for file system modification, network exfiltration, or shell command execution.
  • Sanitization: No sanitization or filtering of the retrieved content is specified before the summary is generated.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 15, 2026, 06:37 AM