linkedin-post-generator

Pass

Audited by Gen Agent Trust Hub on May 9, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [SAFE]: The skill does not contain any evidence of malicious patterns such as hardcoded credentials, data exfiltration, or obfuscation. It manages its secret (COMPOSIO_API_KEY) via environment variables as per standard security practices.
  • [PROMPT_INJECTION]: The skill processes data from external sources (URLs and GitHub PRs), which presents a surface for indirect prompt injection.
  • Ingestion points: In Step 1, the agent fetches content from user-provided URLs or GitHub PR descriptions.
  • Boundary markers: No technical delimiters are specified for the fetched content, but the skill uses strict writing guidelines and negative constraints as logic boundaries.
  • Capability inventory: The skill can execute the linkedin_create_linkedin_post tool to publish content.
  • Sanitization: The skill includes explicit instructions forbidding the invention of metrics or details not found in the source material, acting as a behavioral filter for the output.
  • [EXTERNAL_DOWNLOADS]: The skill fetches source material from external websites and GitHub repositories to fulfill its core function of content summarization.
  • [COMMAND_EXECUTION]: The skill utilizes the Composio integration to perform authorized actions (posting to LinkedIn). The workflow includes a mandatory user confirmation step before calling the tool and handles the API key through secure environment configuration.
Audit Metadata
Risk Level
SAFE
Analyzed
May 9, 2026, 03:23 AM
Security Audit — agent-trust-hub — linkedin-post-generator