Matrix/Credential Harvest/T-3006
T-3006validatedactive13 evidence records

Context Window Credential Leak

Exploit context window contents to leak credentials that were previously processed by the agent

Tactic

Credential Harvest (Stage 3)

Extract API keys, tokens, and credentials from agent context and connected services

Attack Class

RETROACTIVE-PRIV

Exploiting previously granted access or cached credentials to gain unauthorized capabilities

Evidence

validated

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

DVAA Validation

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

LegacyBot credential leak

Honeypot Coverage (AgentPwn)

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

Evidence Source Breakdown

HMA
13 records

Evidence Timeline

HMA

HMA check CRED-001 failed on cred-test

May 12, 2026
HMA

HMA check CRED-001 failed on cred-test

May 11, 2026
HMA

HMA check CRED-001 failed on cred-test

Apr 30, 2026
HMA

HMA check CRED-001 failed on soc-demo

Apr 30, 2026
HMA

HMA check CRED-001 failed on opena2a-cli

Apr 29, 2026
HMA

HMA check CRED-001 failed on cred-test

Apr 29, 2026
HMA

HMA check CRED-001 failed on cred-test

Apr 28, 2026
HMA

HMA check CRED-001 failed on cred-test

Apr 27, 2026
HMA

HMA check CRED-001 failed on cred-test

Apr 22, 2026
HMA

HMA check AGENT-CRED-001 failed on triggerdotdev/trigger.dev

Apr 9, 2026
HMA

HMA check AGENT-CRED-001 failed on @khanglvm/jira-mcp

Apr 9, 2026
HMA

HMA check AGENT-CRED-001 failed on nwiizo/tfmcp

Apr 9, 2026
HMA

HMA check AGENT-CRED-001 failed on saidsurucu/borsa-mcp

Apr 9, 2026

Detection (HackMyAgent)

Live2 live · 0 queued
CRED-001AGENT-CRED-001
npx hackmyagent secure --ciLive = check implemented in hackmyagent; queued = declared, not yet implemented

Defense (OASB Controls)

Live5 live · 0 queued
Live = documented at oasb.ai; queued = declared, not yet documented

How to Cite

AI Agent Threat Matrix T-3006 (Context Window Credential Leak). OpenA2A, 2026. https://threats.opena2a.org/techniques/T-3006