Evidence the run may read
- Approved source and localized asset
- Terminology glossary
- Locale and layout rules
production blueprint · marketing
Localization quality review agent blueprint.
Compare localized assets with approved source meaning, protected terms, layout constraints, and locale guidance.
Search intent: AI agent for localization QA
Published by a2a cloudProduct-source methodology →
01 · the contract
Start from a bounded job, not a blank chat box.
trigger
A localized customer-facing asset enters quality review.
owner
marketing owner
agent stops at
A qualified locale reviewer approves the localized asset for publication.
KPI · decrease
localization defects after publish (defects). Count material meaning, term, or layout issues found after publication.
outputs
Artifacts the run must produce
- Meaning and terminology diff
- Layout-risk checklist
- Reviewer correction proposal
02 · topology
Small specialists. Named handoffs. One accountable decision.
Each stage produces an artifact another stage can inspect. The final node is a person, not an autonomous write to an external system.
- 01
intake
Validate and normalize the localization quality review inputs.
- 02
meaning comparator
Produce the meaning and terminology diff.
- 03
terminology checker
Produce the layout-risk checklist.
- 04
locale reviewer
Challenge the localization quality review result and prepare an approval packet.
- 05
marketing owner
A qualified locale reviewer approves the localized asset for publication.
intake→ validated input packet →meaning-comparator
meaning-comparator→ Meaning and terminology diff →terminology-checker
terminology-checker→ Layout-risk checklist →locale-reviewer
locale-reviewer→ approval packet with evidence references →human-approver
03 · authority
Grant the run only what this case needs.
Source material is read-only. Drafts land in a case-specific output path. Tools may read or propose; the human gate owns the external write.
case inputs
workspace/marketing/localization-quality-review/inputs/**Read only the evidence attached to this workflow instance.
case outputs
workspace/marketing/localization-quality-review/outputs/**Write drafts and evidence artifacts without modifying source records.
approved tools
marketing:localization-quality-review:read-or-proposeInvoke only tools explicitly granted for this run; external writes remain gated.
04 · implementation
A private, bounded starting manifest.
The blueprint starts private, caps its DAG, disables replanning, and exposes no public endpoint. Add only the tools and data adapters this workflow has approved.
a2a.yamlsafe starting point
name: marketing-localization-quality-review
version: 0.1.0
entrypoint: agent:BlueprintAgent
expose:
public: false
composition:
planning: deterministic_dag
max_nodes: 6
max_parallel: 1
max_replans: 005 · acceptance test
Pass only with evidence.
- 01
Evidence traceability
Every material conclusion cites an input artifact or a scoped tool result from this run.
- 02
Critical claims, calls to action, dates, currencies, and protected terms retain approved meaning.
Critical claims, calls to action, dates, currencies, and protected terms retain approved meaning.
- 03
Approval boundary
The run stops at a proposal and records the human decision before any external side effect.
failure containment
Stop small. Preserve the evidence.
The locale, source version, or qualified reviewer is unavailable.
Containment: Return a partial result with unresolved items; do not broaden scope or perform an external write.
Operator: Attach the missing evidence, narrow the brief, or explicitly approve a new scoped run.
A required input or tool grant is unavailable.
Containment: Stop the affected branch and preserve completed artifacts in the case output workspace.
Operator: Grant only the missing resource or continue with that branch marked out of scope.
06 · proof
Sign the run facts. Keep money in the billing ledger.
Current platform receipts sign caller identity or classification, skill, bounded input evidence, verified grant IDs when present, outcome or result preview, and timing. Optional file, tool, artifact, handoff, evaluation, and review fields require separate instrumentation and are not populated by default. Price, fees, payouts, and later human approvals remain separate platform records.
The example uses only fields populated by the current platform sealing paths. It is illustrative, not a record of a real customer run.
ExecutionReceipt · selected fieldsEd25519 token
{
"receipt_id": "rcpt_01J...",
"schema_version": 1,
"agent_name": "marketing-localization-quality-review",
"caller": "user:workflow-owner",
"task_id": "case_localization_quality_review",
"skill_name": "localization_quality_review",
"input_hash": "4d7c...9a2f",
"grant_ids": [
"grt_case_inputs",
"grt_tool_propose"
],
"status": "ok",
"result_preview": "Output prepared: Meaning and terminology diff. Human decision remains separate.",
"elapsed_ms": 4218
}related blueprints
Keep the operating model consistent.
Marketing
Source-backed content brief
Turn an approved topic, audience, and source set into a differentiated brief with claims and evidence mapped explicitly.
Marketing
Campaign quality assurance
Check campaign assets for message consistency, required links, approved claims, and channel-specific readiness.
Marketing
SEO content refresh
Compare a page with current first-party product truth and performance evidence and propose a focused refresh.
put the pattern to work
Start private. Scope the authority. Require the decision.
Deploy the workflow as a bounded internal agent, verify its outputs and the receipt fields actually emitted, then expand only the scopes your acceptance test proves it needs.