github-profile-optimization
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill uses a Python script to fetch metadata and README content from official GitHub API endpoints (api.github.com) and raw content servers (raw.githubusercontent.com). These operations target well-known and expected service domains for the skill's primary function.
- [COMMAND_EXECUTION]: The skill executes a local Python script
scripts/profile-audit.pyto automate profile data collection. This script performs read-only network operations and text analysis on the retrieved content. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted content from external GitHub profile README files. This creates a surface for indirect prompt injection where malicious instructions embedded in a README could attempt to influence the agent's behavior.
- Ingestion points: Retrieval of README content via
scripts/profile-audit.pyor user-provided text. - Boundary markers: No explicit delimiter or 'ignore' instructions are used when processing external README content.
- Capability inventory: The skill can execute Python scripts, access the network (to GitHub), and maintain persistent local state via
devrel-context.md. - Sanitization: No filtering or sanitization of the README content is performed beyond standard markdown parsing and metadata extraction.
- [SAFE]: No malicious patterns such as credential exfiltration, persistence mechanisms, or obfuscation were detected. The use of environment variables for API authentication (GITHUB_TOKEN) is handled according to security best practices.
Audit Metadata