Noisepan Digest

Set up automated news digests using noisepan (signal extraction), entropia (source verification), and HN blind spot detection. Use when configuring daily new...

This skill is an instruction-only integration for noisepan/entropia and appears coherent, but before installing do these simple checks: 1) Inspect the referenced GitHub repos (ppiankov/noisepan and /entropia) and verify they are the projects you expect and have recent trustworthy releases. 2) When using the Linux install, verify the checksums.txt and ideally the release signatures yourself before extracting into /usr/local/bin; prefer a user-local install (~/bin or ~/.local/bin) if you are unsure. 3) Review any helper scripts the SKILL.md writes to ~/.local/bin or /usr/local/bin before making them executable. 4) Ensure you have python3, curl, and sha256sum (or platform equivalents) installed and be aware some commands assume GNU grep (-P). 5) Avoid running install/extract commands as root unless you trust the release artifacts. These steps will reduce risk while allowing the skill to function as described.

Type: OpenClaw Skill Name: noisepan-digest Version: 1.0.5 The skill bundle implements automated news digests but employs several high-risk behaviors, including downloading and installing external binaries from GitHub (ppiankov/noisepan and ppiankov/entropia) and creating shell scripts in SKILL.md that launch a local HTTP server (python3 -m http.server 18222) to bypass rate limits. While these actions are aligned with the stated purpose, the combination of binary execution, local server hosting, and script-based configuration manipulation (using sed and ln) represents a significant attack surface and potential for unintended side effects.