agentmemory-rest-api

Pass

Audited by Gen Agent Trust Hub on Nov 5, 2026

Risk Level: SAFECREDENTIALS_UNSAFEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill instructs the agent to programmatically read a sensitive credential file located at ~/.agentmemory/secret or ~/.agentmemory/.env to authenticate API requests.
  • [DATA_EXFILTRATION]: The instructions establish a pattern of reading a local secret and transmitting it via Authorization headers in curl requests. While the default target is localhost, the skill mentions support for remote daemons, which creates a potential exfiltration vector if the agent is directed to a malicious endpoint.
  • [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to retrieve and store information in a memory server. Content retrieved via /agentmemory/smart-search or /agentmemory/recall is intended to be used by the agent, creating a surface for indirect prompt injection if the stored data contains malicious instructions.
  • Ingestion points: Data enters the agent's context through responses from search and recall endpoints defined in REFERENCE.md.
  • Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions when the agent processes retrieved memory content.
  • Capability inventory: The skill uses shell commands (curl) and local file reads (cat).
  • Sanitization: There is no evidence of content sanitization or validation for data retrieved from the REST API before it is processed by the LLM.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 5, 2026, 05:36 AM