Agent Threat Matrixv1.0
Matrix/Collection/T-7003
T-7003ValidatedActive1 evidence record

API Data Harvesting

Harvest data from connected APIs using the agent's authorized tool access

Tactic

Collection · Stage 7

Gather and stage data from databases, file systems, and APIs

Attack class

SKILL-EXFIL

Using legitimate tool capabilities for unauthorized data transfer

Evidence grade
Validated

Reproduced in a controlled lab environment (DVAA) with documented steps.

DVAA validation

API credential harvesting

Reproductions in Damn Vulnerable AI Agent, the OpenA2A intentionally-broken agent for kill-chain validation.

Honeypot

AgentPwn coverage

Live
data-exfiltrationagentpwn.com/learn ↗

An AgentPwn trap page produces a payload tagged with this technique class. Following the AgentPwn taxonomy of trap pages shows what an agent encounters.

Data-exfiltration tiers harvest data the agent surfaces from connected sources.

Provenance

Evidence by source

Shodan
1 record
Trail

Evidence timeline

Shodan

740 unauthenticated MLflow tracking servers discovered allowing unrestricted API data harvesting

Mar 20, 2026View source
Detect

Detection · HackMyAgent

Live1 live · 3 queued
IO-001IO-002IO-003IO-004
npx hackmyagent secure --ciLive = implemented in hackmyagent; queued = declared
Defend

Defense · OASB controls

Live4 live · 0 queued
OASB 4.1OASB 4.2OASB 4.3OASB 4.4
Live = documented at oasb.ai; queued = declared
Reference

How to cite

AI Agent Threat Matrix T-7003 (API Data Harvesting). OpenA2A, 2026. https://threats.opena2a.org/techniques/T-7003
← Back to the matrix