Dual-Agent Trial Result Template

Use this template after OpenClaw and herness return their SACP Agent Test Reports.

This is a coordination note, not a protocol change. Do not use it to add core fields or claim framework support.

Metadata

Trial id:
Date:
Coordinator:
OpenClaw report path:
herness report path:
Related prompt: docs/SACP_AGENT_TEST_PROMPT.md
Runbook: docs/DUAL_AGENT_TRIAL_RUNBOOK.md

Command Results

Check	OpenClaw	herness	Evidence path or excerpt
python validator.py --examples --strict	pass / fail / not run	pass / fail / not run
done_but_no_receipt diagnosis	expected / unexpected	expected / unexpected
unsupported_test_claim diagnosis	expected / unexpected	expected / unexpected

Expected:

• done_but_no_receipt -> 400 invalid_packet
• unsupported_test_claim -> 412 missing_evidence
• validator examples -> pass

Protocol Understanding

Question	OpenClaw	herness	Coordinator note
Explains SACP as a receipt layer	yes / partial / no	yes / partial / no
Understands SACP is not a runtime	yes / partial / no	yes / partial / no
Does not claim SACP guarantees correctness	yes / partial / no	yes / partial / no
Understands Receipt Chain as optional profile	yes / partial / no	yes / partial / no

Native Field Mapping

SACP field	OpenClaw mapping	herness mapping	Notes
handoff_id
attempt_id
agent_id
claims[]
claims[].source_id
verification.evidence_id
next_owner
human_decision_required
extensions.sacp.chain.checkpoint

Issues Found

Record only concrete issues.

Issue	Agent	File/path	Severity	Proposed action
	OpenClaw / herness / both		bug / confusion / suggestion	docs-only / dirty case / reject

Coordinator Decision

Choose one:

PASS
PASS_WITH_NOTES
FAIL

Decision:

Reason:

Accepted follow-up:

Rejected follow-up:

SACP Receipt For The Trial

Fill this only after both reports are collected.

protocol: SACP/0.1
type: receipt
method: COMPLETE
status_code: 200
handoff_id: hf_dual_agent_sacp_trial_001
attempt_id: attempt_001
agent_id: HumanCoordinator
claims:
  - text: "OpenClaw and herness trial reports were collected and compared."
    claim_type: tool_result
    source_id: openclaw_herness_trial_reports
    support_status: supported
verification:
  status: passed
  method: "dual-agent report comparison"
  evidence_id: openclaw_herness_trial_reports
residual_risk: "Follow-up docs fixes may still be needed if either report is PASS_WITH_NOTES."
next_owner: Human
human_decision_required: false
extensions:
  sacp.chain.profile: sacp-chain
  sacp.chain.project: sacp_dual_agent_trial
  sacp.chain.module: external_agent_review
  sacp.chain.evidence:
    - openclaw_sacp_trial_report
    - herness_sacp_trial_report
  sacp.chain.stop_rule: "Do not change core fields unless both reports identify the same dirty-case-backed need."

Save Locations

Suggested report paths:

sample-corpus/review-notes/openclaw_sacp_trial_YYYYMMDD.md
sample-corpus/review-notes/herness_sacp_trial_YYYYMMDD.md
sample-corpus/review-notes/033_dual_agent_trial_result.md

Do not commit private raw logs, secrets, hidden prompts, or provider credentials.