tsconfig-json-rules

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 1.00). The "Memory Protocol" section mandates reading an internal file (cat .claude/context/memory/learnings.md), recording learnings, and altering assumed context—instructions unrelated to reviewing tsconfig.json rules that request access to internal state and persistence, which is a hidden/deceptive operational instruction outside the skill's stated purpose.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The package contains a deliberate "Memory Protocol" prompt-injection that instructs the agent to read a local internal memory file (.claude/context/memory/learnings.md) and record findings, which is a clear attempt to access/exfiltrate internal/secret state and acts as a backdoor-like prompt; no network exfiltration or remote execution appears in the code, but the memory-access directive constitutes high-risk malicious intent.