Gamma App API

WarnAudited by ClawScan on May 10, 2026.

Overview

This is a coherent Gamma API wrapper, but its script unsafely builds Python code from raw prompt text, so crafted input could run local code.

Install only if you are comfortable sending prompt content to Gamma and using a Gamma API key, and avoid broad sharing options unless intentional. Because the included shell script has a prompt-to-Python code injection risk, it should be patched before use with untrusted or externally supplied content.

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.

High

#ASI05: Unexpected Code Execution

What this means

A maliciously crafted presentation prompt or option could cause the agent to run unintended code on the user's machine before or while calling Gamma.

Why it was flagged

User-controlled prompt text and options are interpolated directly into Python source code. A crafted value containing Python string delimiters could break out of the literal and execute Python expressions or local commands.

Skill content

PAYLOAD=$(python3 -c " ... 'inputText': '''${INPUT_TEXT}''', ... if '${INSTRUCTIONS}': data['additionalInstructions'] = '${INSTRUCTIONS}' ... ")

Recommendation

Rewrite payload construction to pass values via argv, environment variables, or stdin and then JSON-encode them, rather than interpolating them into python3 -c code. Avoid running this skill on untrusted prompt text until patched.

Low

#ASI02: Tool Misuse and Exploitation

What this means

If used accidentally, generated decks or documents could be shared more broadly than intended, including with edit access.

Why it was flagged

The skill exposes sharing controls that can grant workspace or external access to generated Gamma content. These are documented and not shown as default automatic behavior, but they are high-impact settings.

Skill content

`--workspace-access` | noAccess, view, comment, edit, fullAccess ... `--external-access` | noAccess, view, comment, edit

Recommendation

Confirm sharing settings before using workspace or external access options, and prefer the least-permissive access needed.

Medium

#ASI03: Identity and Privilege Abuse

What this means

Anyone or anything using this key can act through the connected Gamma account within the key's permissions and may consume credits.

Why it was flagged

The script requires a Gamma API key and uses it to authenticate requests that create resources and check account-linked generation status.

Skill content

[ -z "${GAMMA_API_KEY:-}" ] && die "GAMMA_API_KEY not set" ... -H "X-API-KEY: ${GAMMA_API_KEY}"

Recommendation

Use the least-privileged Gamma API key available, keep it out of chat/logs, and rotate it if exposed.

Medium

#ASI07: Insecure Inter-Agent Communication

What this means

Sensitive business content included in prompts may be transmitted to Gamma and processed under Gamma's service policies.

Why it was flagged

The artifacts clearly disclose that prompt/content data and the API key are sent to Gamma's external API, which is expected for this integration.

Skill content

The script only makes network calls to `https://public-api.gamma.app/v1.0`. Your API key and inputText content are sent to Gamma's servers.

Recommendation

Do not include confidential or regulated data unless Gamma is approved for that data and the account settings meet your requirements.