charlie-standing-rules
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCREDENTIALS_UNSAFE
Full Analysis
- [SAFE]: The skill consists of high-level engineering principles and process guidelines. It does not contain executable scripts, hidden commands, or malicious instructions.
- [INDIRECT_PROMPT_INJECTION]: The instructions establish a secure workflow for handling external task data from Linear. Ingestion points: Linear ticket descriptions and charter blocks (Rule 33, 34). Boundary markers: Explicit repository routing blocks and mandatory adherence to issue templates (Rule 33, 34). Capability inventory: Standard file and repository operations for coding tasks. Sanitization: Mandatory verification of acceptance criteria and a focused question policy for missing or ambiguous data (Rule 32, 34).
- [CREDENTIALS_UNSAFE]: Rule 11 mandates a secure approach to secret management, explicitly prohibiting the hardcoding or pasting of sensitive values and requiring a formal plan before implementation.
Audit Metadata