invisible-threads

Pass

Audited by Gen Agent Trust Hub on Nov 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests untrusted text data and processes it using an LLM without sufficient isolation.
  • Ingestion points: scripts/chunk_corpus.py reads markdown files from a source directory into a database; scripts/extract_insights.py then reads these segments.
  • Boundary markers: Absent. The EXTRACTION_PROMPT in scripts/extract_insights.py interpolates {chunk_text} directly at the end of the prompt without delimiters or instructions to ignore embedded instructions.
  • Capability inventory: scripts/extract_insights.py makes network calls to Gemini and Claude APIs; scripts/find_threads.py and scripts/extract_insights.py execute shell commands to install packages.
  • Sanitization: Absent. No filtering or escaping is applied to the corpus text before it is sent to the LLM.
  • [COMMAND_EXECUTION]: The scripts use system calls to manage the execution environment.
  • Evidence: scripts/extract_insights.py and scripts/find_threads.py invoke subprocess.check_call() to run pip install commands.
  • [EXTERNAL_DOWNLOADS]: The skill dynamically fetches necessary libraries from public registries during runtime.
  • Evidence: Fetches well-known packages including google-generativeai, python-louvain, and sentence-transformers from PyPI.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 9, 2026, 12:33 PM