Andrew Google Tasks

PassAudited by ClawScan on May 10, 2026.

Overview

The skill is a coherent Google Tasks integration, but authorizing it lets the agent read and change your Google Tasks and stores an OAuth token locally.

Install this only if you are comfortable granting Google Tasks OAuth access to the agent. The provided code appears focused on Google Tasks and does not show hidden exfiltration or background behavior, but you should protect the local token file and use explicit confirmation for task deletion or other changes.

Findings (4)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

After authorization, the agent can access Google Tasks within this scope and the token remains on disk for future use.

Why it was flagged

The skill requires Google OAuth credentials and stores a reusable token for the full Google Tasks scope. This is expected for managing tasks, but it grants account authority.

Skill content
SCOPES = ['https://www.googleapis.com/auth/tasks']
CREDENTIALS_FILE = Path.home() / '.google-credentials.json'
TOKEN_FILE = Path.home() / '.google-tasks-token.pickle'
Recommendation

Only authorize an account you intend to use with this skill, keep the token file private, and revoke the Google OAuth grant if you stop using it.

What this means

If the agent is asked or allowed to act on tasks, it can modify or delete items in the authorized Google Tasks account.

Why it was flagged

The skill exposes API operations that create, update, and delete Google Tasks. These operations match the stated purpose, but they can change user data.

Skill content
created_task = service.tasks().insert(...).execute()
updated_task = service.tasks().update(...).execute()
service.tasks().delete(tasklist=tasklist_id, task=task_id).execute()
Recommendation

Review task-changing requests carefully, especially deletes or bulk changes, and prefer explicit user confirmation for destructive actions.

What this means

The installed package versions may vary depending on when setup is run.

Why it was flagged

The setup instructions install external Python packages without pinned versions. This is normal for a Google API integration, but the exact dependency versions are not fixed in the artifacts.

Skill content
pip install google-api-python-client google-auth-httplib2 google-auth-oauthlib
Recommendation

Install from trusted package indexes and consider pinning or reviewing package versions in controlled environments.

What this means

A malicious local token file could run code when the skill loads it, assuming an attacker can modify that file.

Why it was flagged

The skill deserializes a local pickle token file. This is a common local token-cache pattern, but pickle files can execute code if maliciously replaced or tampered with.

Skill content
with open(TOKEN_FILE, 'rb') as token:
    creds = pickle.load(token)
Recommendation

Do not copy token pickle files from untrusted sources; keep the token file protected and consider safer serialization for future revisions.