Non-technical end users
This test run didn't collect any data because the reports went missing. We can't tell if the AI agents are sharing memories reliably or not. It's like planning a bunch of experiments but losing all the notes.
Campaign a2a-ironclaw-v3r14-mtls-develop FAIL
Infrastructure
Node roster
| # | Role | Agent ID | Public IP | Private IP |
|---|---|---|---|---|
| 1 | agent | ai:alice | 159.203.161.193 | 10.10.2.5 |
| 2 | agent | ai:bob | 45.55.136.89 | 10.10.2.2 |
| 3 | agent | ai:charlie | 104.131.18.124 | 10.10.2.3 |
| 4 | memory-only | — | 104.236.222.25 | 10.10.2.4 |
Baseline attestation BASELINE VIOLATION
Per the authoritative baseline spec, every agent node must emit a self-attestation before any scenario is permitted to run. This run's attestation:
Spec version: 1.0.0 — see authoritative baseline.
| Node | Agent | Framework | Authentic | MCP ai-memory | xAI cfg | xAI default | Agent ID | Federation | UFW off | iptables | dead-man | F1 xAI | F2a substrate | F2b agent (non-gating) | Config SHA | Pass |
|---|
a2a-baseline.json
{
"baseline_pass": false,
"per_node": [],
"failure_mode": "baseline-absent"
}
Run focus
Campaign failed with no scenario reports recovered.
What this campaign tested: The campaign requested 35 scenarios covering AI memory sharing across federation mesh with mTLS, but no reports were recovered, resulting in zero coverage of transport, framework, or primitives axes.
What it demonstrated: The run proved a harness failure in recovering scenario reports, demonstrating nothing about AI memory reliability or functionality.
AI NHI analysis · Claude Opus 4.7
Campaign failed with no scenario reports recovered.
FAIL — no reports recovered from 35 requested scenarios.
For three audiences
C-level decision makers
High risk due to complete campaign failure from report recovery issues, signaling unstable testing infrastructure not ready for production. Customer claims on AI memory sharing remain unvalidated. No progress or changes detectable versus prior runs.
Engineers & architects
Failure mode was total absence of scenario reports, as indicated in summary reasons, with all 35 requested scenarios (S1, S1b, S2, S4-S6, S9-S18, S22-S25, S28-S42) effectively skipped. Probable root cause is a bug in the CI harness's report aggregation or infra teardown phase, possibly related to harness_sha 472a414c9315b4d4455722503b45aa34ed6c2060. No primitives or probes were exercised.
What changes going into the next campaign
Implement logging and error handling in the harness to diagnose and prevent report recovery failures.