thegraph-token-mcp-skill
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data retrieved from a remote API, which could theoretically contain malicious instructions designed to influence the agent's behavior.
- Ingestion points: Data is ingested from the The Graph Token API endpoint at
https://token-api.mcp.thegraph.com/as documented inSKILL.md. - Boundary markers: The skill lacks explicit boundary markers or delimiters to isolate remote API responses from the agent's internal instructions, although it does recommend using JSON output envelopes.
- Capability inventory: The skill uses the
uxctool to execute CLI commands and manage network interactions, as seen inSKILL.mdandreferences/usage-patterns.md. - Sanitization: The skill instructions emphasize parsing stable JSON fields (
ok,kind,protocol,data,error) and performing help-first probing (-h) before execution, which provides a layer of structural validation.
Audit Metadata