Brand & CommsintermediateFree

Build a CEO/CMO alignment diagnostic app

Collect independent leadership views, compare definitions and expectations, visualize shared ground and divergence, facilitate explicit decisions, and publish a jointly approved 30-day operating plan.

What you will have

A private diagnostic app with paired intake, transparent scoring, evidence-linked divergence views, a facilitated workshop, decision records, a 30-day alignment plan, and a follow-up pulse.

Setup time

10-18 hours

Time saved

6-10 hours of interview synthesis and workshop preparation

Estimated cost

$20 to $250 per month

Tools used

4 tools

Why this works

Leadership misalignment often hides inside shared words such as growth, pipeline, brand, and accountability. Independent responses reveal definitions and expectations before negotiation, while deterministic scoring and reviewed qualitative evidence keep the diagnostic inspectable. The app supports a facilitated decision process rather than pretending a score can resolve leadership judgment.

Step-by-step workflow

Run the workflow

This workflow is fully available. Follow the steps below to build the system from start to finish.

Define the alignment dimensions and confidentiality rules

60-90 min

Define dimensions such as company strategy, growth model, marketing mandate, time horizon, pipeline, brand, customer, positioning, budget, risk, metrics, decision rights, talent, and CEO-CMO working cadence. Write operational definitions and examples for each dimension. Decide which responses remain private until the joint session and who can view raw answers. Establish that the diagnostic supports a conversation and is not a performance evaluation. Record the operation against stable identifiers such as diagnostic_version, respondent_role, dimension_id, question_id, raw_response, preserve the raw source reference and capture time, and write any transformation or decision into the system’s change history rather than replacing the prior value. Before the step is marked complete, the facilitator validates scoring and privacy while the CEO and CMO approve only the shared decisions and action plan, not each other’s private raw notes; if that check fails, suppress low-confidence coding, prevent role-identifying data leakage, and present scoring disagreements as discussion prompts rather than objective diagnoses.

Output

An approved diagnostic model and confidentiality agreement.

Airtable

Pro tip

Psychological safety determines answer quality. Raw responses should not be forwarded to a broader leadership team without explicit consent.

Create separate CEO and CMO intake instruments

2-3 hours

Build two Tally forms with parallel questions but role-appropriate wording. Capture priorities, definitions, expected outcomes, time horizons, confidence, perceived constraints, decision rights, current tensions, and one example per major answer. Use scales only when the endpoints are concrete. Add free-text questions for what marketing should stop, start, protect, and decide. Record the operation against stable identifiers such as respondent_role, dimension_id, question_id, raw_response, coded_value, preserve the raw source reference and capture time, and write any transformation or decision into the system’s change history rather than replacing the prior value. Before the step is marked complete, the facilitator validates scoring and privacy while the CEO and CMO approve only the shared decisions and action plan, not each other’s private raw notes; if that check fails, suppress low-confidence coding, prevent role-identifying data leakage, and present scoring disagreements as discussion prompts rather than objective diagnoses.

Output

Paired CEO and CMO questionnaires with comparable and role-specific fields.

TallyAirtable

Pro tip

Ask both leaders to define pipeline, brand, and growth in their own words. Agreement on a numeric score can hide different definitions.

Map fields and calculate transparent agreement measures

1-2 hours

Create an Airtable field map linking paired questions to dimensions, weights, response types, and scoring logic. Use deterministic formulas for numeric gaps and documented coding rules for categorical or text responses. Keep a raw-response view separate from the joint summary. Do not let an LLM generate the final alignment score without inspectable inputs. Record the operation against stable identifiers such as dimension_id, question_id, raw_response, coded_value, evidence_excerpt, preserve the raw source reference and capture time, and write any transformation or decision into the system’s change history rather than replacing the prior value. Before the step is marked complete, the facilitator validates scoring and privacy while the CEO and CMO approve only the shared decisions and action plan, not each other’s private raw notes; if that check fails, suppress low-confidence coding, prevent role-identifying data leakage, and present scoring disagreements as discussion prompts rather than objective diagnoses.

Output

A transparent scoring model with raw, coded, and summary layers.

Airtable

Pro tip

The overall score should never replace the dimension view. Averages can hide one severe disagreement on budget or mandate.

Collect responses independently

20-30 min per leader

Send separate private links and ask each leader to complete the diagnostic without coordinating answers. Record completion, version, and timestamp. Do not disclose one leader’s response to the other before both are complete. Route incomplete or contradictory form records to the facilitator for clarification. Record the operation against stable identifiers such as question_id, raw_response, coded_value, evidence_excerpt, agreement_score, preserve the raw source reference and capture time, and write any transformation or decision into the system’s change history rather than replacing the prior value. Before the step is marked complete, the facilitator validates scoring and privacy while the CEO and CMO approve only the shared decisions and action plan, not each other’s private raw notes; if that check fails, suppress low-confidence coding, prevent role-identifying data leakage, and present scoring disagreements as discussion prompts rather than objective diagnoses.

Output

Independent CEO and CMO response sets under the same diagnostic version.

TallyAirtable

Pro tip

Independent completion reveals real assumptions. A jointly completed form often produces negotiated answers before the underlying difference is visible.

Code open responses with evidence excerpts

30-45 min

Use Claude to map open-text answers to the defined dimensions, definitions, expectations, tensions, and decision-right themes. Preserve short evidence excerpts and do not infer emotion, intent, or relationship quality. Identify materially different word meanings and questions that need clarification. Have the facilitator review all coded differences. Run this template in the workflow's dedicated Claude Project after attaching the named inputs, and save the structured response to the step's governed review record. Record the operation against stable identifiers such as raw_response, coded_value, evidence_excerpt, agreement_score, divergence_type, preserve the raw source reference and capture time, and write any transformation or decision into the system’s change history rather than replacing the prior value. Use an explicit pass, warning, or hold disposition, attach the supporting evidence IDs, and assign every unresolved exception to an owner and due date before moving to the next step.

Output

Reviewed qualitative coding linked to exact response evidence.

ClaudeAirtable

Pro tip

Language differences can matter more than score differences. `Marketing owns pipeline` may mean lead volume to one leader and revenue accountability to the other.

Prompt template

ROLE
You are coding confidential CEO and CMO diagnostic responses for a facilitated alignment session.

OBJECTIVE
Map the text to predefined dimensions and surface definition, expectation, constraint, and decision-right differences without judging either person.

INPUTS
Diagnostic dimensions and definitions: {{dimensions}}
CEO responses: {{ceo_responses}}
CMO responses: {{cmo_responses}}
Scoring and confidentiality rules: {{rules}}

WORK TO PERFORM
Code each response, preserve short evidence excerpts, identify shared views, divergent definitions, expectation gaps, constraint gaps, and questions requiring clarification.

OUTPUT SCHEMA
Return `coded_responses`, `shared_ground`, `definition_gaps`, `expectation_gaps`, `constraint_gaps`, `decision_right_gaps`, `clarification_questions`, and `do_not_infer`.

GUARDRAILS
Do not infer personality, competence, emotion, blame, or relationship health. Do not expose one leader’s private answer beyond the approved joint-summary rules.

EVIDENCE REQUIREMENTS
Tie every factual conclusion, score, or recommendation to the supplied input field, record ID, source ID, timestamp, or approved rule that supports it. Do not convert an unsupported assumption into a fact. When the supplied evidence is insufficient or conflicting, mark the item `unverified` and route it through the uncertainty and human-review sections.

UNCERTAINTY HANDLING
Use confidence per code and explain ambiguous language. Keep uncertain items as facilitator questions.

HUMAN REVIEW
The facilitator reviews all coding and decides what may appear in the joint report.

Generate the divergence and shared-ground map

30-45 min

Combine deterministic score gaps with reviewed qualitative codes. Show each dimension’s agreement, uncertainty, definition mismatch, business impact, urgency, and evidence. Separate healthy difference from unresolved conflict and missing information. Avoid one simplistic alignment grade; use a profile with the highest-impact divergences and strongest shared ground. Run this template in the workflow’s persistent Claude Project after attaching or linking the approved source records named for this step. Record the operation against stable identifiers such as coded_value, evidence_excerpt, agreement_score, divergence_type, decision_id, preserve the raw source reference and capture time, and write any transformation or decision into the system’s change history rather than replacing the prior value. Before the step is marked complete, the facilitator validates scoring and privacy while the CEO and CMO approve only the shared decisions and action plan, not each other’s private raw notes; if that check fails, suppress low-confidence coding, prevent role-identifying data leakage, and present scoring disagreements as discussion prompts rather than objective diagnoses.

Output

A dimension-level alignment profile with evidence and business impact.

ClaudeAirtable

Pro tip

Two leaders can disagree productively on tactics while being aligned on mandate and outcomes. The map should not pathologize every difference.

Prompt template

ROLE
You are the governed analysis and operations assistant supporting the facilitator, CEO, and CMO. You are working inside the CEO/CMO alignment diagnostic application, where traceability, stable identifiers, and human authority matter more than producing a polished but unsupported answer.

OBJECTIVE
Complete workflow step 6, “Generate the divergence and shared-ground map,” and produce this operational outcome: A dimension-level alignment profile with evidence and business impact. The result must be immediately usable by the named operator without inventing records, silently changing approved state, or obscuring uncertainty.

INPUTS
1. SOURCE RECORDS: {{generate_the_divergence_and_shared_ground_map_source_records}}
2. FIELD DICTIONARY AND ALLOWED VALUES: {{generate_the_divergence_and_shared_ground_map_field_dictionary}}
3. OPERATING, PERMISSION, AND DECISION RULES: {{generate_the_divergence_and_shared_ground_map_operating_rules}}
4. APPROVAL CONTEXT, OWNERS, AND DEADLINES: {{generate_the_divergence_and_shared_ground_map_approval_context}}
5. PRIOR VERSION, SNAPSHOT, OR CURRENT STATE: {{generate_the_divergence_and_shared_ground_map_prior_version_or_state}}
Authoritative evidence may include independent Tally responses, scoring rules, coded excerpts, workshop decisions, and follow-up pulse results.

WORK TO PERFORM
1. Execute the specific job described by “Generate the divergence and shared-ground map”; do not broaden the task into a generic strategy exercise.
2. Use the canonical field names and IDs supplied in the inputs, especially diagnostic_version, respondent_role, dimension_id, question_id, raw_response, coded_value.
3. Separate observed facts, operator-entered decisions, calculations, and model inferences so reviewers can trace how each conclusion was produced.
4. Return records that can be copied into the CEO/CMO alignment diagnostic application without renaming identifiers or collapsing one-to-many relationships.
5. Preserve raw values, source identifiers, capture timestamps, transformation notes, and exception status so the operation is reproducible and reversible.
6. Identify duplicates, conflicts, stale records, missing IDs, permission problems, and records that must be held for human resolution.
7. Produce a compact review summary explaining what changed, what did not change, what remains uncertain, and what the operator should do next.

OUTPUT SCHEMA
Return valid JSON only, using this exact top-level structure:
{
"workflow_slug": "ceo-cmo-alignment-diagnostic-app",
"step_number": 6,
"step_title": "Generate the divergence and shared-ground map",
"run_status": "pass|warning|hold|fail",
"source_records": [
{"source_id": "string", "source_type": "string", "captured_at": "ISO-8601|null", "authoritative": true, "notes": "string|null"}
],
"records": [
{"diagnostic_version": "value|null", "respondent_role": "value|null", "dimension_id": "value|null", "question_id": "value|null", "raw_response": "value|null", "coded_value": "value|null", "evidence_excerpt": "value|null", "evidence_source_ids": ["string"], "confidence": "high|medium|low", "review_status": "approved|needs-review|held"}
],
"exceptions": [
{"record_id": "string|null", "exception_type": "string", "severity": "low|medium|high|critical", "evidence": "string", "owner": "string", "required_action": "string"}
],
"changes_from_prior_state": [
{"record_id": "string", "field": "string", "prior_value": "value|null", "proposed_value": "value|null", "reason": "string", "source_ids": ["string"]}
],
"review_summary": {"facts": ["string"], "inferences": ["string"], "open_questions": ["string"], "next_actions": [{"action": "string", "owner": "string", "due_date": "YYYY-MM-DD|null"}]},
"qa": {"schema_valid": true, "ids_preserved": true, "evidence_complete": true, "human_approval_required": true}
}

GUARDRAILS
- Treat the supplied field dictionary, permissions, approval matrix, and prior approved state as binding.
- Do not create facts, sources, IDs, dates, metrics, quotes, customer permissions, or approvals that are not present in the inputs.
- Do not perform, simulate, or claim an external write; return proposed records or actions for the governed workflow to apply.
- Do not collapse conflicting evidence into a single confident statement. Preserve the conflict and identify the required owner.
- suppress low-confidence coding, prevent role-identifying data leakage, and present scoring disagreements as discussion prompts rather than objective diagnoses.

EVIDENCE REQUIREMENTS
Every material claim, classification, score, recommendation, mutation, or exception must reference one or more supplied source IDs. Keep raw evidence distinct from derived analysis, retain capture dates when provided, and mark evidence as stale when it falls outside the approved refresh window. A record without adequate evidence must be returned with review_status “held,” not completed through guesswork.

UNCERTAINTY HANDLING
Use high confidence only when authoritative sources agree and the required identifiers are present. Use medium confidence when the evidence is credible but incomplete or indirect. Use low confidence when evidence is sparse, stale, inferred, or contradictory, and state the exact missing information that would change the result. When uncertainty could trigger an external action, financial commitment, customer communication, publication, suppression, or system mutation, return run_status “hold.”

HUMAN REVIEW
The facilitator, CEO, and CMO must review the JSON before any state change or external action. The approval gate is: the facilitator validates scoring and privacy while the CEO and CMO approve only the shared decisions and action plan, not each other’s private raw notes. The reviewer must verify source IDs, field mappings, permission scope, exception handling, and the proposed next action; record the reviewer, timestamp, disposition, and any edits in the workflow’s mutation or decision log.

Prototype the diagnostic experience as a Claude Artifact

2-4 hours

Create an interactive Artifact using synthetic or anonymized data. Include completion status, shared-ground view, divergence heatmap, definition comparison, decision-right matrix, and facilitator questions. Keep raw private answers out of the shared artifact unless both leaders approved them. Test print, screen-share, and mobile views for workshop use. Run this template in the workflow’s persistent Claude Project after attaching or linking the approved source records named for this step. Record the operation against stable identifiers such as evidence_excerpt, agreement_score, divergence_type, decision_id, owner, preserve the raw source reference and capture time, and write any transformation or decision into the system’s change history rather than replacing the prior value. Use an explicit pass, warning, or hold disposition, attach the supporting evidence IDs, and assign every unresolved exception to an owner and due date before moving to the next step.

Output

A reviewable interactive prototype for the alignment workshop.

Claude

Pro tip

Use the artifact to make the conversation concrete, not to create a theatrical score. The best interaction highlights the next decision.

Prompt template

OBJECTIVE
Complete workflow step 7, “Prototype the diagnostic experience as a Claude Artifact,” and produce this operational outcome: A reviewable interactive prototype for the alignment workshop. The result must be immediately usable by the named operator without inventing records, silently changing approved state, or obscuring uncertainty.

INPUTS
1. SOURCE RECORDS: {{prototype_the_diagnostic_experience_as_a_claude__source_records}}
2. FIELD DICTIONARY AND ALLOWED VALUES: {{prototype_the_diagnostic_experience_as_a_claude__field_dictionary}}
3. OPERATING, PERMISSION, AND DECISION RULES: {{prototype_the_diagnostic_experience_as_a_claude__operating_rules}}
4. APPROVAL CONTEXT, OWNERS, AND DEADLINES: {{prototype_the_diagnostic_experience_as_a_claude__approval_context}}
5. PRIOR VERSION, SNAPSHOT, OR CURRENT STATE: {{prototype_the_diagnostic_experience_as_a_claude__prior_version_or_state}}
Authoritative evidence may include independent Tally responses, scoring rules, coded excerpts, workshop decisions, and follow-up pulse results.

WORK TO PERFORM
1. Execute the specific job described by “Prototype the diagnostic experience as a Claude Artifact”; do not broaden the task into a generic strategy exercise.
2. Use the canonical field names and IDs supplied in the inputs, especially diagnostic_version, respondent_role, dimension_id, question_id, raw_response, coded_value.
3. Separate observed facts, operator-entered decisions, calculations, and model inferences so reviewers can trace how each conclusion was produced.
4. Return records that can be copied into the CEO/CMO alignment diagnostic application without renaming identifiers or collapsing one-to-many relationships.
5. Create the requested deliverable from approved evidence only, preserve citations or source IDs beside every material claim, and identify what must remain internal.
6. Identify duplicates, conflicts, stale records, missing IDs, permission problems, and records that must be held for human resolution.
7. Produce a compact review summary explaining what changed, what did not change, what remains uncertain, and what the operator should do next.

OUTPUT SCHEMA
Return valid JSON only, using this exact top-level structure:
{
"workflow_slug": "ceo-cmo-alignment-diagnostic-app",
"step_number": 7,
"step_title": "Prototype the diagnostic experience as a Claude Artifact",
"run_status": "pass|warning|hold|fail",
"source_records": [
{"source_id": "string", "source_type": "string", "captured_at": "ISO-8601|null", "authoritative": true, "notes": "string|null"}
],
"records": [
{"diagnostic_version": "value|null", "respondent_role": "value|null", "dimension_id": "value|null", "question_id": "value|null", "raw_response": "value|null", "coded_value": "value|null", "evidence_excerpt": "value|null", "evidence_source_ids": ["string"], "confidence": "high|medium|low", "review_status": "approved|needs-review|held"}
],
"exceptions": [
{"record_id": "string|null", "exception_type": "string", "severity": "low|medium|high|critical", "evidence": "string", "owner": "string", "required_action": "string"}
],
"changes_from_prior_state": [
{"record_id": "string", "field": "string", "prior_value": "value|null", "proposed_value": "value|null", "reason": "string", "source_ids": ["string"]}
],
"review_summary": {"facts": ["string"], "inferences": ["string"], "open_questions": ["string"], "next_actions": [{"action": "string", "owner": "string", "due_date": "YYYY-MM-DD|null"}]},
"qa": {"schema_valid": true, "ids_preserved": true, "evidence_complete": true, "human_approval_required": true}
}

Build the deployable app in Lovable

6-10 hours

Implement private role-based access, Tally or Airtable data ingestion, deterministic scoring, reviewed qualitative summaries, the shared report, and session notes. Keep secrets in environment variables and store only the minimum required response data. Add an export for the final decision plan and a deletion or archival process. Do not expose one leader’s raw answers through client-side code. Record the operation against stable identifiers such as diagnostic_version, respondent_role, dimension_id, question_id, raw_response, preserve the raw source reference and capture time, and write any transformation or decision into the system’s change history rather than replacing the prior value. Before the step is marked complete, the facilitator validates scoring and privacy while the CEO and CMO approve only the shared decisions and action plan, not each other’s private raw notes; if that check fails, suppress low-confidence coding, prevent role-identifying data leakage, and present scoring disagreements as discussion prompts rather than objective diagnoses.

Output

A deployable private diagnostic app with controlled views and exports.

LovableAirtableTally

Pro tip

Test access with separate CEO, CMO, and facilitator accounts. Role leakage is a critical release blocker.

Run privacy, scoring, and edge-case QA

2-3 hours

Test incomplete responses, changed answers, version mismatch, tied scores, extreme divergence, missing examples, access denial, and export. Recalculate sample scores independently and verify qualitative evidence links. Review data retention, sharing, deletion, and facilitator permissions. Record blockers and rollback instructions. Record the operation against stable identifiers such as respondent_role, dimension_id, question_id, raw_response, coded_value, preserve the raw source reference and capture time, and write any transformation or decision into the system’s change history rather than replacing the prior value. Use an explicit pass, warning, or hold disposition, attach the supporting evidence IDs, and assign every unresolved exception to an owner and due date before moving to the next step.

Output

A signed QA matrix for logic, access, privacy, and output accuracy.

LovableAirtable

Pro tip

The diagnostic may contain politically sensitive leadership views. Security and access QA matter as much as scoring accuracy.

Facilitate the alignment workshop

90-120 min

Begin with shared ground, then review the highest-impact definition and expectation gaps. For each gap, distinguish fact, assumption, preference, constraint, and decision. Record one agreed definition, decision owner, success measure, next action, and review date. Do not use the app to declare a winner; use it to structure explicit decisions. Record the operation against stable identifiers such as dimension_id, question_id, raw_response, coded_value, evidence_excerpt, preserve the raw source reference and capture time, and write any transformation or decision into the system’s change history rather than replacing the prior value. Use an explicit pass, warning, or hold disposition, attach the supporting evidence IDs, and assign every unresolved exception to an owner and due date before moving to the next step.

Output

A facilitated decision record and aligned operating definitions.

LovableAirtable

Pro tip

Resolve language before targets. Agreeing what marketing-sourced and marketing-influenced mean can unlock several downstream metric decisions.

Publish the 30-day alignment plan

30-45 min

Use Claude to convert workshop decisions into a concise plan with mandate, priorities, definitions, metrics, decision rights, meeting cadence, stop-doing list, actions, owners, dates, and unresolved items. Link every plan item to a workshop decision rather than generating new recommendations. Have both leaders approve the plan. Store the approved version and share only the agreed view. Run this template in the workflow's dedicated Claude Project after attaching the named inputs, and save the structured response to the step's governed review record. Record the operation against stable identifiers such as question_id, raw_response, coded_value, evidence_excerpt, agreement_score, preserve the raw source reference and capture time, and write any transformation or decision into the system’s change history rather than replacing the prior value. Use an explicit pass, warning, or hold disposition, attach the supporting evidence IDs, and assign every unresolved exception to an owner and due date before moving to the next step.

Output

A jointly approved 30-day CEO-CMO alignment plan.

ClaudeAirtable

Pro tip

Keep unresolved items explicit. False consensus creates more damage than a clearly owned disagreement with a review date.

Prompt template

ROLE
You are documenting the output of a facilitated CEO-CMO alignment workshop.

OBJECTIVE
Convert agreed decisions into an operating plan without adding new strategy or smoothing unresolved disagreement.

INPUTS
Approved workshop decisions: {{workshop_decisions}}
Agreed definitions: {{definitions}}
Decision-right matrix: {{decision_rights}}
Actions and owners: {{actions}}
Unresolved items: {{unresolved_items}}
Review cadence: {{review_cadence}}

WORK TO PERFORM
Organize the decisions into a practical 30-day plan and preserve exact ownership, dates, metrics, and unresolved questions.

OUTPUT SCHEMA
Return `marketing_mandate`, `shared_priorities`, `agreed_definitions`, `success_measures`, `decision_rights`, `stop_start_protect`, `actions`, `unresolved_items`, and `review_schedule`.

GUARDRAILS
Do not invent consensus, change owners, introduce new recommendations, or omit unresolved tensions. Keep private raw responses out of the plan.

EVIDENCE REQUIREMENTS
Tie every factual conclusion, score, or recommendation to the supplied input field, record ID, source ID, timestamp, or approved rule that supports it. Do not convert an unsupported assumption into a fact. When the supplied evidence is insufficient or conflicting, mark the item `unverified` and route it through the uncertainty and human-review sections.

UNCERTAINTY HANDLING
Flag decisions with ambiguous wording or missing dates and return them for facilitator correction.

HUMAN REVIEW
Both CEO and CMO approve the final plan before it is distributed.

Recheck alignment and maintain the diagnostic

45-60 min per pulse

At 30 or 60 days, repeat a short pulse on the highest-impact dimensions and review action completion, metric use, decision speed, and recurring disagreements. Compare against the original diagnostic version and note company-context changes. Update questions, weights, examples, and app versions only after facilitator review. Archive or delete raw responses according to the confidentiality agreement. Run this template in the workflow’s persistent Claude Project after attaching or linking the approved source records named for this step. Record the operation against stable identifiers such as raw_response, coded_value, evidence_excerpt, agreement_score, divergence_type, preserve the raw source reference and capture time, and write any transformation or decision into the system’s change history rather than replacing the prior value. Use an explicit pass, warning, or hold disposition, attach the supporting evidence IDs, and assign every unresolved exception to an owner and due date before moving to the next step.

Output

A follow-up alignment pulse and maintained diagnostic version.

TallyAirtableClaude

Pro tip

A better score is not the only success. Faster decisions, fewer definition debates, and completed commitments may be stronger evidence of alignment.

Prompt template

OBJECTIVE
Complete workflow step 12, “Recheck alignment and maintain the diagnostic,” and produce this operational outcome: A follow-up alignment pulse and maintained diagnostic version. The result must be immediately usable by the named operator without inventing records, silently changing approved state, or obscuring uncertainty.

INPUTS
1. SOURCE RECORDS: {{recheck_alignment_and_maintain_the_diagnostic_source_records}}
2. FIELD DICTIONARY AND ALLOWED VALUES: {{recheck_alignment_and_maintain_the_diagnostic_field_dictionary}}
3. OPERATING, PERMISSION, AND DECISION RULES: {{recheck_alignment_and_maintain_the_diagnostic_operating_rules}}
4. APPROVAL CONTEXT, OWNERS, AND DEADLINES: {{recheck_alignment_and_maintain_the_diagnostic_approval_context}}
5. PRIOR VERSION, SNAPSHOT, OR CURRENT STATE: {{recheck_alignment_and_maintain_the_diagnostic_prior_version_or_state}}
Authoritative evidence may include independent Tally responses, scoring rules, coded excerpts, workshop decisions, and follow-up pulse results.

WORK TO PERFORM
1. Execute the specific job described by “Recheck alignment and maintain the diagnostic”; do not broaden the task into a generic strategy exercise.
2. Use the canonical field names and IDs supplied in the inputs, especially diagnostic_version, respondent_role, dimension_id, question_id, raw_response, coded_value.
3. Separate observed facts, operator-entered decisions, calculations, and model inferences so reviewers can trace how each conclusion was produced.
4. Return records that can be copied into the CEO/CMO alignment diagnostic application without renaming identifiers or collapsing one-to-many relationships.
5. Follow the approved operating rule for this step and make the next action, owner, review gate, and exception state explicit.
6. Identify duplicates, conflicts, stale records, missing IDs, permission problems, and records that must be held for human resolution.
7. Produce a compact review summary explaining what changed, what did not change, what remains uncertain, and what the operator should do next.

OUTPUT SCHEMA
Return valid JSON only, using this exact top-level structure:
{
"workflow_slug": "ceo-cmo-alignment-diagnostic-app",
"step_number": 12,
"step_title": "Recheck alignment and maintain the diagnostic",
"run_status": "pass|warning|hold|fail",
"source_records": [
{"source_id": "string", "source_type": "string", "captured_at": "ISO-8601|null", "authoritative": true, "notes": "string|null"}
],
"records": [
{"diagnostic_version": "value|null", "respondent_role": "value|null", "dimension_id": "value|null", "question_id": "value|null", "raw_response": "value|null", "coded_value": "value|null", "evidence_excerpt": "value|null", "evidence_source_ids": ["string"], "confidence": "high|medium|low", "review_status": "approved|needs-review|held"}
],
"exceptions": [
{"record_id": "string|null", "exception_type": "string", "severity": "low|medium|high|critical", "evidence": "string", "owner": "string", "required_action": "string"}
],
"changes_from_prior_state": [
{"record_id": "string", "field": "string", "prior_value": "value|null", "proposed_value": "value|null", "reason": "string", "source_ids": ["string"]}
],
"review_summary": {"facts": ["string"], "inferences": ["string"], "open_questions": ["string"], "next_actions": [{"action": "string", "owner": "string", "due_date": "YYYY-MM-DD|null"}]},
"qa": {"schema_valid": true, "ids_preserved": true, "evidence_complete": true, "human_approval_required": true}
}

Expected results

Preparation time

6-10 hours saved

Paired forms, field maps, deterministic calculations, coding prompts, and app views replace manual interview synthesis.

Alignment visibility

Dimension-level shared ground and divergence

Definitions, expectations, constraints, decision rights, and confidence remain visible instead of collapsing into one score.

Workshop output

Named decisions, owners, and dates

The facilitated process converts differences into explicit operating agreements and unresolved items.

Follow-through

30-day plan plus pulse

Leaders approve a practical plan and recheck the highest-impact dimensions after execution.

Related workflows

Continue with workflows that share a similar GTM motion, category, or tool stack.

Brand & Commspro

Build an executive POV intake and evidence desk

A persistent executive editorial desk that selects defensible ideas, verifies claims, captures first-hand credibility, and hands approved POV briefs to downstream production.

ClaudeGranolaPerplexity

5-8 hoursView workflow ->

ABMpro

Build a one-account interactive ABM experience

A production-ready interactive account experience with approved narrative, evidence, inputs, branches, results, PostHog instrumentation, QA, privacy controls, and account-team action.

ClayClaudeLovable

16-28 hoursView workflow ->

Marketing Opspro

Maintain a client current-state and decision log agent

A persistent, evidence-backed client operating system with current state, decisions, owners, unresolved issues, weekly briefs, and an auditable change history.

ClaudeNotionGranola

6-10 hoursView workflow ->