social-clips

Pass

Audited by Gen Agent Trust Hub on Apr 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to perform several shell-based operations, including dependency installation with npm, media rendering using npx remotion and ffmpeg, and asset retrieval via curl.
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to download user avatar images from Slack's official CDN (avatars.slack-edge.com). These downloads are restricted to image assets required for the video rendering process.
  • [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection due to its core functionality of processing external data.
  • Ingestion points: Data enters the agent's context through the mcp__slack__slack_get_thread_replies tool call specified in SKILL.md.
  • Boundary markers: No specific delimiters or instructions to ignore embedded commands within the fetched Slack messages are provided.
  • Capability inventory: The environment allows for subprocess execution (remotion, ffmpeg), network access (curl), and file system modifications (writing TypeScript data files and deleting temporary assets).
  • Sanitization: The implementation in src/components/slack/SlackMessageRow.tsx focuses on visual formatting (bolding, mentions) and lacks mechanisms to sanitize the raw text against potential instructions that could influence the agent during the data preparation phase.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 4, 2026, 06:40 AM