google-calendar
2 findings (1 high, 1 medium), mainly data exfiltration, social engineering. Most recent scan 2026-02-06.
This Google Calendar skill presents SIGNIFICANT SECURITY CONCERNS primarily related to credential management and third-party trust boundaries. The skill's architecture routes all Google Calendar API requests and OAuth tokens through a third-party gateway (Maton), creating substantial data exposure risks. Users must trust Maton with full access to their Google Calendar data and OAuth credentials. The documentation lacks critical security disclosures about this architecture and encourages insecure credential handling practices (API keys in command-line arguments and environment variables). While the skill does not contain malicious code (no scripts provided), its design inherently creates high-risk trust dependencies that are not adequately disclosed to users. The missing referenced file (Google.py) and absent metadata fields further reduce transparency. This skill should only be used if users explicitly trust Maton with their Google Calendar access and understand the security implications.
Scanner's overall assessment (LLM analyzer)
Second scanner: NVIDIA SkillSpector
Independent static run, no language model, over the same files as the Cisco scan above. How the two scanners compare across the whole corpus: Two scanners, one corpus.
What this skill does
Google Calendar API integration with managed OAuth. Create events, list calendars, check availability, and manage schedules. Use this skill when users want to interact with Google Calendar.
- Package
- 2 files, 8.9 KB
- SKILL.md
- 669 words, 12+ sections
- Code
- None. Instructions only.
- Compatibility
- Requires network access and valid Maton API key
SKILL.md covers: Google Calendar · Quick Start · List upcoming events · Base URL · Authentication · Getting Your API Key · Connection Management · List Connections · Create Connection · Get Connection · Delete Connection · Specifying Connection
Findings
HIGH Third-Party API Gateway Credential Delegation Data Exfiltrationllm
The skill routes all Google Calendar API requests through a third-party gateway (gateway.maton.ai) that 'automatically injects your OAuth token'. This creates a significant trust boundary issue where the Maton gateway has full access to users' Google Calendar OAuth tokens and can intercept, log, or exfiltrate all calendar data passing through it. Users must trust Maton with their Google credentials and calendar data.
SKILL.md
Base URL: https://gateway.maton.ai/google-calendar/{native-api-path}
'The gateway proxies requests to www.googleapis.com and automatically injects your OAuth token.'MEDIUM Missing Critical Security Disclosures Social Engineeringllm
The skill description claims 'managed OAuth' as a feature but does not disclose that this involves routing all API requests and OAuth tokens through a third-party service (Maton). Users are not informed about the security implications of this architecture, including data exposure to Maton's infrastructure, potential logging, or the trust requirements. This constitutes deceptive omission of material security information.
SKILL.md
Description: 'Google Calendar API integration with managed OAuth. Create events, list calendars, check availability, and manage schedules.'
No disclosure that 'managed OAuth' means Maton gateway has access to OAuth tokens and all calendar data.What is in the package
| File | Kind | Size |
|---|---|---|
LICENSE.txt | doc | 1.0 KB |
SKILL.md | doc | 7.8 KB |
External hosts referenced
developers.google.com gateway.maton.ai ctrl.maton.ai maton.ai connect.maton.ai
Hostnames that appear in the skill's files. Referencing a host is not a finding; compare against the findings above and the skill's stated purpose.
Environment variables referenced
MATON_API_KEY
Package contents recorded 2026-10-22. The scan above ran against the version current at scan time; the two can differ if the skill has since been updated.
What we checked
- Static analyzer. Pattern rules over SKILL.md and bundled files: known-bad commands, exfiltration URLs, secrets, obfuscation.
- Behavioral analyzer. Checks what the skill asks the agent to do against what its description claims it does.
- Llm analyzer. A language model reads the skill and reports intent-level issues the rules cannot express.
- Trigger analyzer. Looks for instructions that fire on conditions the user did not ask for.
- Meta analyzer. A second model pass that validates, deduplicates and re-scores the other analyzers' findings.
What we did not check
- Runtime behaviour. The skill was not executed; findings come from reading its files.
- Anything fetched at install or run time, including remote scripts, package dependencies, and URLs the skill tells the agent to open.
- Content the analyzers cannot read: images, binaries, and some non-English text.
- Known scanner bypass classes. See the methodology and limitations page.
Sources
- Repository: https://clawhub.ai/skills/google-calendar-api
- Raw scan result: google-calendar-scan.json
- Badge for your README:
- Request a re-scan