The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes instructions for the agent to execute shell commands using curl and python3. These commands are used to fetch staking data from the network and perform basic calculations on the JSON output.
[EXTERNAL_DOWNLOADS]: The skill performs network requests to official Injective blockchain endpoints (e.g., injective.network). These are vendor-owned resources used to retrieve staking, reward, and validator information.
[DYNAMIC_EXECUTION]: Example activities in SKILL.md utilize the python3 -c flag to execute a static, embedded script for processing JSON data from the API. This is a local execution pattern for data analysis.
[INDIRECT_PROMPT_INJECTION]: Because the skill processes data from a public blockchain (such as validator names and descriptions), it is susceptible to indirect prompt injection. A malicious actor could populate blockchain metadata with instructions designed to influence the agent's behavior when it reads that data.
Ingestion points: Data retrieved from https://lcd.injective.network and https://tm.injective.network via REST and RPC calls.
Boundary markers: The skill does not provide specific delimiters or instructions to the agent to treat blockchain data as untrusted.
Capability inventory: The skill possesses capabilities for network communication (API calls), subprocess execution (curl, python3), and blockchain interaction via the pyinjective SDK.
Sanitization: No explicit sanitization or filtering of user-controllable blockchain fields is documented in the skill instructions.

injective-trading-staking