fable-security
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest and analyze untrusted data such as source code, diffs, and API responses, which creates a surface for indirect prompt injection attacks where instructions could be embedded in the audited data.
- Ingestion points: The skill processes code, diffs, and security-sensitive files as defined in its activation criteria (SKILL.md).
- Boundary markers: The skill includes instructions to identify and reason about trust boundaries, but does not provide explicit technical delimiters for separating instructions from processed data.
- Capability inventory: The skill defines high-level reasoning protocols but does not contain executable scripts or tool-call definitions that could be abused for privilege escalation or network access.
- Sanitization: A dedicated protocol for secret sanitization and masking is provided to handle credentials detected during audits (references/secret-sanitization.md).
- [SAFE]: The skill contains regular expression patterns for identifying common API keys, tokens, and private keys. These patterns are used for detection and sanitization purposes within its security auditing workflow and do not include any actual hardcoded credentials.
Audit Metadata