findall

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSNO_CODE
Full Analysis
  • Unverifiable Dependencies (MEDIUM): The skill instructs the user to install an external package findall-cli via npm (npm install -g findall-cli). This package is not from a trusted organization or repository defined in the security policy, posing a risk of supply chain attack or execution of unvetted code.
  • Indirect Prompt Injection (LOW): The skill accepts untrusted natural language input through the <objective> argument which is then processed by a "generator" (LLM tier).
  • Ingestion points: The <objective> string passed to the findall run command in SKILL.md.
  • Boundary markers: Absent. There are no instructions provided to the agent or the tool to delimit the user objective or ignore potentially malicious instructions embedded within the search criteria.
  • Capability inventory: The CLI tool performs entity discovery and writes results to the local file system (matches.json, matches.csv, etc.).
  • Sanitization: Absent. No evidence of input validation or instruction filtering is present in the skill definition.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 06:22 PM