handoff

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt's HTTP fallback explicitly instructs including an Authorization: Bearer $AGENTMEMORY_SECRET header, which requires the agent/LLM to insert a secret API token verbatim into generated requests/output, creating an exfiltration risk.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill falls back to making runtime HTTP requests to $AGENTMEMORY_URL/agentmemory/sessions and $AGENTMEMORY_URL/agentmemory/recall which fetch remote observations that are injected into the agent's context and therefore can directly control prompts/behavior.