save-note

Pass

Audited by Gen Agent Trust Hub on Oct 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use the system date utility (specifically date -u +%Y%m%dT%H%M%SZ) to generate UTC timestamps for file naming and record entries within the SKILL.md instructions.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection because it reads previously saved records from local storage to resume conversation runs.
  • Ingestion points: Reads files from the plugin store at ${CLAUDE_PLUGIN_DATA}/records/ or the ${user_config.records_mirror} directory as described in SKILL.md.
  • Boundary markers: Records are structured using JSONL format, which helps delimit data fields.
  • Capability inventory: The skill uses system command execution (date), file reading, and file writing to manage note records.
  • Sanitization: No explicit sanitization or escaping of the content read from historical record files is specified.
  • [EXTERNAL_DOWNLOADS]: The README.md points to installation via npx from the author's GitHub repository and references example records and license information hosted on the same domain (github.com/synapseradio/ai-skills). These references originate from the skill author's infrastructure.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 9, 2026, 09:00 AM