Silviu Core

Implements Silviu-specific guards and runbooks to enforce validation, prevent common errors, and manage browser and GitHub interactions in OpenClaw.

This is an instruction-only runbook bundle that appears coherent and low-risk: it checks browser relay status and performs harmless GitHub preflight commands (curl to github.com, git ls-remote) and suggests exporting GIT_TERMINAL_PROMPT to avoid interactive prompts. Before enabling, confirm you trust the unknown publisher; although the runbooks are small and transparent, instruction-only skills execute shell commands and will make network requests (to github.com) and run git operations, which are visible in logs and to network monitoring. If you have stricter network or audit requirements, review the exact commands and run them manually first. No credentials are requested by the skill.

Type: OpenClaw Skill Name: silviu-core Version: 1.0.0 The skill bundle includes instructions for the AI agent to execute shell commands like `curl -I https://github.com`, `export GIT_TERMINAL_PROMPT=0`, and `git ls-remote <repo>` as part of a 'Repo Audit Preflight' runbook in `SKILL.md` and `repo_audit_preflight.md`. While these commands are plausibly needed for the stated purpose, the execution of arbitrary shell commands, especially with a placeholder like `<repo>` that could be substituted by the agent, represents a risky capability and a potential command injection vulnerability if the agent's execution environment does not properly sanitize input. This falls under 'risky capabilities without clear malicious intent'.