odyssey4me/google-slides
Create and edit Google Slides presentations. Add or delete slides, insert text, shapes, and images. Use when asked to build a deck, create a slideshow, update a Google presentation, or edit slides.
75
94%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
View guide
High
Do not use without reviewing
Security
1 high severity finding. You should review these findings carefully before considering using this skill.
W007: Insecure credential handling detected in skill instructions.
What this means
The skill handles credentials insecurely by requiring the agent to include secret values verbatim in its generated output. This exposes credentials in the agent’s context and conversation history, creating a risk of data exfiltration.
Why it was flagged
The prompt shows examples that embed OAuth client_id/client_secret in a config file and as direct CLI flags (--client-secret), which would require the agent to accept and output secret values verbatim (high exfiltration risk).
Where we found it
[REDACTED]
secret · 1 site
The SKILL.md documentation shows a config file example with a placeholder client_secret value that an agent would substitute with a real OAuth client secret, creating a credential emission site.
SKILL.md
50
client_secret: [REDACTED]
--client-secret
command · 7 sites
The CLI accepts --client-secret as a required argument for auth setup, which would require the agent to pass a real OAuth client secret on the command line, exposing it in process listings and shell history.
Low
Low-risk findings.
1 low severity finding. Worth noting, but not necessarily harmful.
W011: Third-party content exposure detected (indirect prompt injection risk).
What this means
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
Why it was flagged
The skill explicitly reads and ingests user/third-party slide content via the presentations read command and read_presentation_content function (scripts/google-slides.py) and accepts arbitrary public image URLs via images create (--image-url), so untrusted, user-generated content from external presentations or URLs can be fetched and influence subsequent actions.
Where we found it
docs.google.com
domain · 11 sites
The presentations read command and read_presentation_content function fetch and ingest slide content from Google Slides presentations (hosted at docs.google.com), which can contain arbitrary user/third-party content that could influence subsequent agent actions.
--image-url (arbitrary user-provided image URLs)
url · 7 sites
The images create command accepts arbitrary --image-url values that are passed to the Google Slides API, which fetches the image content from the provided URL, allowing untrusted third-party image content to be ingested.
- Audited
- Security analysis