recap

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The fallback explicitly instructs constructing HTTP requests with an Authorization: Bearer $AGENTMEMORY_SECRET header, which requires the agent to embed a secret token verbatim into outgoing requests/commands.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill falls back at runtime to calling $AGENTMEMORY_URL (GET $AGENTMEMORY_URL/agentmemory/sessions and POST $AGENTMEMORY_URL/agentmemory/recall) and injects the returned observations into the agent's context to produce the recap, so external content from that URL can directly control the agent's prompts/output.