developer-community
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- PROMPT_INJECTION (SAFE): No malicious injection patterns such as 'ignore previous instructions' or bypasses were detected. The persona and grounding instructions in SKILL.md are standard for role-based AI agents.
- DATA_EXFILTRATION (SAFE): The skill does not include any commands for network access (curl, wget) or sensitive file system operations. No hardcoded credentials or secrets were found in the metadata or reference files.
- REMOTE_CODE_EXECUTION (SAFE): There are no scripts, dynamic execution patterns, or remote downloads. The skill is entirely composed of static markdown documentation.
- INDIRECT_PROMPT_INJECTION (SAFE): 1. Ingestion points: User queries regarding developer community management. 2. Boundary markers: Instructions in SKILL.md specify grounded responses using local reference files. 3. Capability inventory: The skill has no capability for file writing, network calls, or subprocess execution. 4. Sanitization: No sanitization logic is present as the skill performs no high-risk actions.
- OBFUSCATION (SAFE): No Base64, zero-width characters, or hidden unicode markers were found. All files are in readable markdown format.
Audit Metadata