Skills from clawhub.ai/davida-ps

This skill provides a security advisory feed for OpenClaw and includes installation and update scripts. It instructs running shell commands (`curl`, `unzip`, `shasum`) and fetching from remote URLs (`https://api.github.com`, `https://raw.githubusercontent.com`), so agents will perform network access and command execution during install.

Clawtributor installs an opt-in incident-reporting tool that writes reports and state under `~/.openclaw` and submits reports to GitHub (`https://api.github.com/repos/prompt-security/ClawSec/releases`, `https://github.com/prompt-security/ClawSec`). The skill contains explicit shell installation commands (`curl`, `unzip`, `gh`) and reads `CLAWTRIBUTOR_INSTALL_DIR`/`$HOME` paths.

Soul-guardian detects drift and maintains tamper-evident baselines and an audit log for workspace files like `SOUL.md` and `AGENTS.md`. It instructs running `python3 skills/soul-guardian/scripts/soul_guardian.py` and performs atomic restores that overwrite workspace files (`atomic_write_bytes(abs_path, read_bytes(approved_snap))`).

This skill schedules daily `openclaw` security audits, renders reports, and delivers them by DM/email using the included scripts (`scripts/runner.sh`, `scripts/run_audit_and_format.sh`). It runs shell commands, reads env vars such as `PROMPTSEC_DM_CHANNEL` and `PROMPTSEC_DM_TO`, and performs network operations (e.g., `curl "https://github.com/prompt-security/clawsec/releases/download/$VERSION_TAG/openclaw-audit-watchdog.skill"` and SMTP to `127.0.0.1`).

This skill manages advisory-feed monitoring, affected-skill detection, and approval-gated installs using `https://raw.githubusercontent.com/prompt-security/clawsec/main/advisories/feed.json` and local state `~/.openclaw/clawsec-suite-feed-state.json`. It invokes system commands like `npx` and `openclaw` and performs network fetches to the listed feed/checksum URLs.