The Agent Skills Directory

[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests untrusted social media data (KOL sentiment, influencer trends, and community chatter) to influence its economic hypotheses and automated decision-making.
Ingestion points: Data enters the context via social intelligence endpoints like https://x402.ottoai.services/kol-sentiment as described in SKILL.md and references/catalogue-findings.md.
Boundary markers: The skill documentation does not define specific delimiters or instructions to ignore embedded commands within the fetched social data.
Capability inventory: The agent possesses the capability to provision mailboxes, manage Circle wallets, and execute gated financial transactions using endpoints defined in references/endpoints.md.
Sanitization: No explicit sanitization or filtering logic is provided for the external social signals.
[EXTERNAL_DOWNLOADS]: The skill fetches data and manages operational identity through several remote services.
Retrieves market context and sentiment analysis from ottoai.services endpoints.
Manages operational email and retrieves OTPs for wallet authentication via agentmail.to.
Performs infrastructure availability checks through stabledomains.dev.
These domains represent infrastructure within the vendor's ecosystem.

self-evolving-agent