mantle-data-indexer

Pass

Audited by Gen Agent Trust Hub on Nov 8, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the mantle-cli tool to perform indexing tasks. It constructs shell commands using runtime-provided parameters such as endpoint URLs and query strings for SQL or GraphQL.- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes data from external indexer endpoints. This creates an attack surface where maliciously crafted data from a compromised or hostile indexer could attempt to influence the agent's logic.
  • Ingestion points: Results from mantle-cli indexer subgraph and mantle-cli indexer sql (SKILL.md, references/indexer-selection-and-quality.md).
  • Boundary markers: The skill employs a structured 'Mantle Historical Data Report' to delineate metadata and query results (SKILL.md).
  • Capability inventory: The agent executes CLI commands with network access to user-specified endpoints (SKILL.md).
  • Sanitization: The instructions mandate that SQL queries be read-only and that all data outputs include source attribution and unit normalization to maintain integrity (SKILL.md, references/query-templates.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 8, 2026, 08:20 PM