Parallel 1.0.1

High-accuracy web search and research via Parallel.ai API. Optimized for AI agents with rich excerpts and citations.

This skill appears to do what it says (wrap the Parallel.ai search API) but contains two practical inconsistencies you should care about: (1) both scripts provide a hardcoded API key fallback — treat that as suspicious (it could be a leaked or shared credential that enables tracking, billing, or unauthorized access). (2) The shell script uses curl and jq and the Python script requires the 'parallel' SDK, yet the registry declares no required binaries/env vars. Before installing: ask the author to remove the embedded API key and to document required binaries and environment variables. If you still want to use it, set your own PARALLEL_API_KEY in your environment and verify the key's permissions and billing implications; audit network traffic to api.parallel.ai and review Parallel.ai's privacy/policy. If you cannot validate the origin of the embedded key or the author's identity, do not install or run these scripts with sensitive queries or on a system where unauthorized API usage would be a problem.

Type: OpenClaw Skill Name: parallel-1-0-1 Version: 1.0.0 The skill bundle is benign. It provides wrappers for the Parallel.ai API for web search and research, as stated in the `SKILL.md` description. The `SKILL.md` contains no prompt injection attempts. Both `scripts/parallel.sh` and `scripts/search.py` interact solely with the `api.parallel.ai` endpoint, using a default API key (which can be overridden by an environment variable) and passing user input as data within JSON payloads, preventing command injection. There is no evidence of data exfiltration, persistence mechanisms, or other malicious activities.