skills/letta-ai/skills/social-cli/Gen Agent Trust Hub

social-cli

Pass

Audited by Gen Agent Trust Hub on Apr 23, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill requires cloning the social-cli source code from github.com/letta-ai/social-cli.git and installing dependencies using pnpm install. These resources are provided by the author of the skill.
  • [COMMAND_EXECUTION]: The skill operates by executing command-line instructions such as social-cli sync, social-cli check, and social-cli dispatch to interact with social media APIs and local YAML state files.
  • [PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by design, as it fetches untrusted data from social media platforms intended for agent processing.
  • Ingestion points: Untrusted notifications and feed data are pulled into inbox-*.yaml and feed.yaml files via the sync and feed commands.
  • Boundary markers: The documentation does not specify the use of delimiters or safety instructions when passing fetched social media text to an AI agent.
  • Capability inventory: The agent possesses capabilities to post, reply, follow, and delete content on social media, which could be exploited via malicious content in the inbox.
  • Sanitization: The documentation does not describe any sanitization or validation of the fetched social media content before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 23, 2026, 06:38 AM