The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources including CoinGecko, DexScreener, Polymarket APIs, and WebSearch results to calculate signals and conviction tiers. This data is processed and used to generate notifications and logs without explicit boundary markers or sanitization, potentially allowing malicious content in the source data to influence the agent's behavior or output.
Ingestion points: External API responses from api.coingecko.com, api.dexscreener.com, and gamma-api.polymarket.com, as well as WebSearch results (Steps 1, 2, 3, and 4).
Boundary markers: Absent for external data ingestion.
Capability inventory: Shell command execution (curl), file system writes (memory/logs), and notification dispatch (./notify).
Sanitization: No sanitization or validation of the content of the API responses or search results is performed before interpolation into the agent's decision logic and output.
[COMMAND_EXECUTION]: The skill uses curl within shell command blocks to retrieve market data from external APIs. While these operations are standard for the skill's defined purpose, direct shell invocation represents a persistent attack surface.
[EXTERNAL_DOWNLOADS]: The skill fetches data from several external services including CoinGecko, DexScreener, and Polymarket. These are well-known services within the cryptocurrency ecosystem, and the skill's data retrieval is consistent with its primary purpose of market analysis.

token-pick