skills/acta-team/skills/acta/Gen Agent Trust Hub

acta

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill handles untrusted data in the form of Verifiable Credential payloads (vcData).\n
  • Ingestion points: Input fields in SKILL.md, references/api.md, and SDK hooks in examples/react-quickstart.tsx.\n
  • Boundary markers: The data is handled as structured JSON conforming to W3C standards, which mitigates the risk of instructions being executed within the data stream.\n
  • Capability inventory: The skill performs API calls to the ACTA REST service and interacts with Stellar wallets for transaction signing; it does not possess arbitrary system execution capabilities.\n
  • Sanitization: Input is validated by the SDK and the backend API during issuance and verification processes.\n- [COMMAND_EXECUTION]: Documented shell examples in examples/api-curl.sh utilize python -c for JSON processing. This is a standard utility usage within a local context and is considered safe.\n- [EXTERNAL_DOWNLOADS]: The skill uses official packages (@acta-team/credentials) and communicates with the author's primary infrastructure (acta.build). These are vendor-owned resources necessary for the skill's function.\n- [CREDENTIALS_UNSAFE]: The skill explicitly promotes secure secret management, advising against hardcoding API keys and providing methods to resolve keys from environment variables.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 04:28 AM
Security Audit — agent-trust-hub — acta