Obsidian Tasknotes

Manage tasks in Obsidian via TaskNotes plugin API. Use when user wants to create tasks, list tasks, query by status or project, update task status, delete tasks, or check what they need to do.

This skill appears to be a legitimate local TaskNotes client, but take a few precautions before installing: - Verify the registry metadata: the package claims no required env vars but SKILL.md and the script expect TASKNOTES_API_PORT and TASKNOTES_API_KEY in a .env — update or confirm the metadata if you control the registry entry. - Inspect where the skill will be stored and check what VAULT_ROOT resolves to in the script (SCRIPT_DIR.parent.parent.parent.parent). Ensure it will point to your Obsidian vault and not to another directory that contains a .env with unrelated secrets (AWS keys, tokens, etc.). - If you don't want any token stored, enable TaskNotes HTTP API with no auth (SKILL.md mentions leaving token empty) and avoid creating a .env. - Review the included script yourself (scripts/tasks.py) and confirm it only talks to http://localhost:<port> and does not send data externally. The script appears to only contact localhost and print results. - Run in a restricted environment or sandbox if you are unsure, and avoid putting high-value secrets in a .env file that could be picked up by this script. If these checks look good, the skill's behavior is coherent with its purpose. If you cannot verify the VAULT_ROOT path or the .env contents, do not install or run it with sensitive environment files present.

Type: OpenClaw Skill Name: obsidian-plugin-tasknotes Version: 0.1.0 The OpenClaw AgentSkills skill bundle for Obsidian TaskNotes is benign. It is designed to interact with a local HTTP API running within Obsidian on `localhost` (default port 8080). The `SKILL.md` provides clear instructions for setup and usage, without any prompt injection attempts. The `scripts/tasks.py` script uses standard Python libraries (`requests`, `python-dotenv`), loads API credentials from a local `.env` file, and makes all network calls exclusively to `localhost`. There is no evidence of data exfiltration, malicious execution, persistence mechanisms, or obfuscation.