Skip to main content
Glama

Continue workflow draft

continue_workflow_draft
Destructive

Resumes a READY AD_STRUCTURE_CREATE draft from persisted server state. This call may query Amazon Ads for keyword or ASIN candidates and suggested bids, including candidates not yet created in the account, or reuse saved recommendations. It persists workflow progress and decisions. It reconciles the last atomic plan, retains successful Amazon IDs, and prepares exactly the next Campaign, Ad Group, Product Ad, automatic-target bid, or Targeting plan. All Campaigns are prepared together, then all Ad Groups, then all Product Ads across every ad group in one multi-group plan; the Product Ad plan is completed before any targeting recommendation or bid decision can pause another structure. It never confirms or applies a write plan. For AUTO ad groups it persists the selected bid mode with the created adGroupId, synchronizes the four Amazon-created targets, obtains suggested bids when selected, and returns a separately confirmable AD_TARGETING_CHANGE plan. Partial writes retry only failed items. A 429 suggestion read includes per-target evidence; ambiguous create results require reconciliation and are never retried. For MANUAL ad groups without supplied positive targeting, it waits until Product Ads exist, persists Amazon candidates only for the recorded KEYWORD or ASIN TARGET family, accepts customer selection by candidateRef, obtains suggested or custom bids, accepts explicit customCandidates when recommendations are absent or rejected, and prepares separately confirmable Targeting plans. It accepts at most 100 selected candidates per ad group and reuses the exact suggested bids already shown to the customer. Once every manual decision is ready, it combines all MANUAL positive targeting groups into one batch plan, then combines all configured negative targeting across AUTO and MANUAL groups into one separate batch plan. AUTO-only drafts skip MANUAL processing and MANUAL-only drafts skip AUTO processing. When a plan is returned, pass applyContinuation unchanged to the matching operation-specific execution tool so a successful write automatically returns the next stage preview without asking the customer to advance. After an AUTO bid write, immediately present the returned MANUAL targeting preview or selection action in the same response when manual groups remain.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
draftIdYesREADY AD_STRUCTURE_CREATE draft ID.
expectedDraftVersionYesLatest draftVersion used for optimistic locking and safe recovery.
manualTargetingDecisionsNoOptional decisions keyed by a deferred MANUAL ad-group clientRef. First submit up to 100 candidateRef values returned by MANUAL_TARGETING_SELECTION_REQUIRED, or add customer-approved customCandidates when recommendations are absent or rejected. All candidates for one ad group must be KEYWORD or TARGET, never both; use separate manual Campaign/Ad Group structures when both targeting families are required. Then choose AMAZON_SUGGESTED, or CUSTOM with one bid for every selected candidateRef. Selection and bid choice may be submitted together when the customer already requested suggested bids.
automaticTargetBidDecisionsNoOptional decisions keyed by AUTO ad-group clientRef, used after a DECIDE_LATER or incomplete recommendation result.

Output Schema

TableJSON Schema
NameRequiredDescriptionDefault
planNo
draftNo
nextActionNo
applyContinuationNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observed

TDQS

A4.5/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Beyond the annotations, the description discloses substantial behavioral details: it queries Amazon Ads, persists workflow progress and decisions, reconciles the last atomic plan, retains successful IDs, and prepares only the next atomic plan. It also describes partial-write retry behavior, 429 suggestion-read evidence, ambiguous create-result reconciliation, and the rule that ambiguous creates are never retried. The destructive/open-world annotations are not contradicted; this tool mutates draft/workflow state while explicitly not applying Amazon Ads writes.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The purpose is front-loaded in the first sentence, which is good. However, the description is a very long, dense paragraph covering many edge cases, sequencing rules, and retry semantics without internal structure. Much of the content is substantive for this complex workflow, but the overall presentation is over-verbose rather than concise.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's orchestration complexity and the presence of an output schema, the description is comprehensive. It explains the workflow stages, AUTO vs MANUAL branching, retry/reconciliation behavior, candidate limits, and how to continue into execution tools. An agent has enough context to invoke and route this tool correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the schema already documents all four parameters and their nested structures. The description still adds useful workflow-level meaning: candidateRef-based selection, customCandidates as a fallback, the 100-candidate cap, one targeting family per ad group, and bid-mode behavior for AUTO and MANUAL decisions. It stops short of fully explaining every parameter interaction because the schema itself is already very detailed.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb and resource: it resumes a READY AD_STRUCTURE_CREATE draft from persisted server state. It also explicitly distinguishes this orchestration/preparation tool from execution tools by saying it never confirms or applies a write plan. The scope and sequencing are detailed enough that an agent can identify this as the draft-continuation tool rather than a get, prepare, or apply sibling.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The precondition is explicit: the draft must be READY and AD_STRUCTURE_CREATE. The description also gives strong routing guidance for the next step by telling the agent to pass applyContinuation unchanged to the matching execution tool. It does not explicitly name when to use get_workflow_draft or prepare_workflow_draft instead, so it falls short of full when/when-not/alternatives guidance.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources