Scan nearby BLE MAC addresses to mint unique soulbound $ANIMA tokens stored locally using hash and DAG-based proof-of-remnant logic.

Before installing or running this skill, consider: - Function vs claim: The code only scans BLE MAC addresses, hashes them with a fixed SALT, and stores hashes locally — it does NOT contact a blockchain, perform any real 'minting', or implement gossip sync despite those claims. If you expect true token minting, this skill is incomplete or misleading. - Privacy: It collects nearby device MACs (hashed). A static SALT means hashes are linkable across runs and could be used to track devices; if you need privacy, avoid a fixed salt or avoid collecting MACs at all. - Serialization risk: It writes anima_dag.gpickle (pickle-based). Unpickling gpickle files from untrusted sources can execute arbitrary code. Do not share or load that file without review; prefer safer formats (JSON) if you will reuse data. - System access: aioblescan requires access to BLE hardware and may need elevated permissions or specific platform setup. Run in an isolated environment (VM/container) and ensure you understand required privileges. - If you plan to extend this to real minting or gossip, require clear network endpoints, authentication keys, and audit of any code that transmits data. Right now there is no external transmission, so immediate exfiltration risk is low, but the mismatch between description and code is suspicious. Recommendations: run only in an isolated test environment, review or replace the fixed SALT, change serialization to a safe format, and do not rely on this code for any real blockchain minting without implementing and auditing secure network and credential handling.

Type: OpenClaw Skill Name: ble-anima-minter Version: 1.0.0 The skill bundle is classified as benign. The `anima_minter.py` script performs BLE scanning to detect MAC addresses, hashes them with a hardcoded salt, and stores the resulting hashes in a local DAG file (`anima_dag.gpickle`) using the `networkx` library. While BLE scanning requires elevated privileges (e.g., raw socket access), this is the core stated purpose of the skill. There is no evidence of data exfiltration, malicious execution, persistence mechanisms, or prompt injection attempts against the agent in `SKILL.md`. All actions are directly aligned with the described functionality.