The Agent Skills Directory

[SAFE]: The skill interacts with the Launkr API (launkr.io) and Hiro API (api.hiro.so) for protocol metadata and blockchain interactions, which is consistent with its stated purpose of managing DeFi operations on the Stacks network.
[SAFE]: A significant security control is implemented via the verifyDeploySourceMatchesTemplate function, which fetches the reference template contract source directly from the Stacks blockchain (via Hiro API) and performs a byte-by-byte comparison against the source provided by the Launkr API. This prevents the agent from deploying modified, malicious, or unverified contract code.
[SAFE]: The skill performs rigorous validation of transaction parameters through the validatePoolStepMatchesRequest function. It ensures that the pool-creation arguments returned by the backend API—specifically the fee receiver address, token supply, and AMM curve parameters—exactly match the user's requested configuration, protecting against potential backend compromises or data manipulation.
[SAFE]: All write operations (swaps and pool creation) use Stacks post-conditions in Deny mode. These are cryptographically signed assertions that limit asset movement, ensuring that only the specified amount of STX or tokens can leave the user's wallet, effectively mitigating risk from unexpected contract behaviors.
[SAFE]: The skill includes a dedicated recovery path (create-pool) for multi-step operations like token launches, allowing the agent to resume an interrupted process without re-spending gas on duplicate deployments.

launkr