debugging-log-analyser
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill instructions focus on technical root-cause analysis and do not contain any patterns indicative of prompt injection, privilege escalation, or persistence.- [SAFE]: Privacy safeguards are integrated into the workflow, specifically instructing the agent to redact secrets from logs to prevent data exposure.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of raw logs and stack traces provided by the user.
- Ingestion points: The 'Required Inputs' section in SKILL.md explicitly asks for raw log, stack trace, and error output content.
- Boundary markers: The skill lacks specific boundary markers or 'ignore' instructions for the data being analyzed, creating a potential surface for instructions embedded within logs to influence the agent.
- Capability inventory: The skill does not request any permissions via 'allowed-tools' and performs no file writes or network operations, effectively neutralizing the risk of data exfiltration or system modification.
- Sanitization: The agent is instructed to redact sensitive information before providing output, though no specific filtering for prompt injection markers is included.
Audit Metadata