CCTV Guardian for Muse

Prerequisites: An EZVIZ account (email + password) with at least one camera added in the EZVIZ app. Python 3.9+. No GitHub account needed unless listed here. Install "CCTV Guardian for Muse" for me. It connects my Muse agent to EZVIZ cloud cameras for snapshots, PTZ moves, and recordings, plus a scheduled motion-alert guardian. Repository: https://github.com/titodoni/cctv-guardian-muse 1) Fetch: clone the repository into a temp folder (or download the files). Read SKILL.md and ONBOARDING.md, then summarize what this build does in 1-2 sentences. 2) Integrity check: Skill Harbor reviewed this listing on 2026-10-05 and recorded file fingerprints. Compute the SHA-256 of each downloaded file and compare with: (Aucune empreinte de fichier enregistrée pour cette fiche — le contrôle d'intégrité est ignoré.) If any fingerprint does NOT match: STOP without installing. Tell me "These files changed since Skill Harbor's review, so the new version is unverified", show your safety findings, and offer three choices: (a) wait for Skill Harbor to re-review the listing (recommended), (b) analyze the new files in detail first and then decide, (c) install anyway knowing the risk. Do nothing further until I answer. 3) Safety check: examine the code and SHOW me explicit findings, never silently pass. List every network address the code contacts (EZVIZ cloud API region endpoints; NOTIFY_WEBHOOK and VISION_API_URL only if I configure them — never hardcoded); every shell command it runs (scripts invoking each other via subprocess; pip install for requirements); every place credentials, API keys, or personal data are read, stored, or sent ($CCTV_GUARDIAN_HOME/.env at mode 600; session and media-key caches at mode 600; sent only to EZVIZ's own login API and to endpoints I configure myself); and any obfuscated or encoded code (there is none — plain readable Python throughout). If anything looks suspicious, stop and explain it before continuing. 4) Confirmation: do NOT install yet. Give me a short summary of what the code does plus the safety findings, and wait for my explicit "yes, install". Proceed only on approval. 5) Install (skill): copy SKILL.md, ONBOARDING.md, and the scripts/ folder into my agent's skills directory (e.g. ~/.muse/skills/cctv-guardian/, or the platform equivalent). Run pip install -r requirements.txt. Then follow ONBOARDING.md with me conversationally: collect my EZVIZ credentials through secure entry (never in plain chat), auto-discover my cameras, take a test snapshot and show it to me, optionally set up my notification webhook and AI descriptions, and optionally create the guardian schedule. 6) Verify with no network calls: SKILL.md frontmatter is valid and every script passes a Python syntax check. 7) Report: what was installed, where it lives, and anything left for me (e.g. adding a webhook URL later, enabling the guardian schedule). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.