Agent Skill · Anthropic
PowerPoint (pptx) skill
Anthropic's published skill for generating and editing PowerPoint decks with correct layouts, placeholders and slide structure.
This skill bundles executable code. Scripts shipped alongside the instructions run with the permissions of the engineer running the agent. Review the scripts, not only the SKILL.md.
What it is, and what it costs you
Slide generation is a layout problem masquerading as a content problem. Getting text into the right placeholder, respecting a template's masters, and producing a deck that does not need manual repair afterwards requires operating on the presentation's object model rather than describing slides in prose. This skill provides the code path for that.
The exposure profile is closest to the Word skill, with one addition worth calling out: decks are the format most likely to be shared outside the organisation, and the most likely to contain material that was never meant to be. Speaker notes are a routine hiding place for candid commentary, figures, and internal context. Hidden slides survive in the file. A deck assembled or edited by an agent from source material carries whatever came along.
Templates are the other consideration. A corporate template received as a starting point is a file with embedded structure, and a template sourced from outside the organisation is untrusted input in the same way any other received document is.
The practical risk for most teams is not attack but leakage: an agent producing a customer-facing deck from internal source material, and nobody checking the notes pane before it is sent.
Capability profile
What this skill can reach inside your environment, and why a security team cares. Derived from the publisher's documentation, not from a scan.
- bundled script execution
Executes with the invoking user's permissions against presentation files.
- speaker notes and hidden slides
Content invisible in presentation mode persists in the file and travels with it when shared externally.
- deck generation from source material
Internal source content can propagate into an externally-shared artifact without a review step.
- template ingestion
Templates are structured files; externally-sourced ones are untrusted input.
Before you distribute this to a team
- Review notes and hidden slides before sharing
The likeliest real incident with this skill is leakage through the notes pane, not compromise.
- Keep generation source material scoped
Constrain which directories the agent draws from so internal material does not reach a customer deck by accident.
- Review and pin the bundled scripts
Read the code, pin the version, re-review on change.
Vincosha Assay assessment
not yet assessedWe have not published a scan verdict for PowerPoint (pptx) skill.
Everything above is a capability profile compiled by hand from the publisher's own public documentation on 2026-07-25. It describes what this artifact can reach and what to restrict before rolling it out. It is not a security verdict, and it should not be read as one.
When Vincosha Assaypublishes an assessment for this artifact it will appear here with a verdict, the findings behind it, and the date it ran. We would rather leave this space empty than fill it with a guess — inventing a security verdict about somebody else's software would be both wrong and dangerous.
What an assessment covers
- prompt-injection patterns in descriptions and instructions
- credential reads and outbound egress in bundled code
- over-scoped permissions and manifest claims
- typosquatted and impersonating names
- tool-manifest drift since the last approval
- content hash against the version you approved
Sources
Compiled from the publisher's own documentation and repository. Skills change — read the version you are about to install rather than trusting a profile.