Skills
skills/c0ntr0lledcha0s/claude-code-plugin-automations/reading-logseq-data/Gen Agent Trust Hub

reading-logseq-data

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTIONREMOTE_CODE_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to Indirect Prompt Injection. It retrieves data from a user's Logseq graph which could contain malicious instructions. Ingestion points: Methods like get_page and search in scripts/logseq-client.py read data from the graph. Boundary markers: No delimiters are used to separate graph content from agent instructions. Capability inventory: The skill uses subprocess.run and the agent has access to Bash and Read tools. Sanitization: Retrieved content is not sanitized.
  • [COMMAND_EXECUTION]: The scripts/logseq-client.py script executes CLI commands using subprocess.run to interact with the Logseq tool.
  • [DATA_EXFILTRATION]: The skill manages sensitive API tokens for Logseq authentication, which could be compromised if the API endpoint is redirected.
  • [REMOTE_CODE_EXECUTION]: The search method in scripts/logseq-client.py constructs Datalog queries using unsafe string interpolation, which is vulnerable to query injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 09:58 AM