palace-index-curator

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local management scripts (scripts/memory_palace_cli.py) using uv run to generate corpus reports and apply index updates. These operations involve reading the index, creating backups in data/backups/, and modifying local files.
  • [PROMPT_INJECTION]: The skill establishes an indirect prompt injection surface by surfacing content from external web captures into the agent's active session.
  • Ingestion points: Content is read from hooks/memory-palace-index.yaml, which aggregates data from external sources via WebFetch and WebSearch hooks.
  • Boundary markers: The instructions do not define specific delimiters or isolation mechanisms for the surfaced content in the index_surfacer.py hook.
  • Capability inventory: The agent can execute commands via uv run and perform file writes for index updates and backups.
  • Sanitization: There is no documented sanitization or validation of the captured web content before it is surfaced to the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 01:39 AM
Security Audit — agent-trust-hub — palace-index-curator