Install
$ agentstack add skill-stephenrogan-augmented-csm-cc-qbr-deck-builder ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
QBR Deck Builder
Generates data-populated QBR decks by assembling usage metrics, health trends, support data, value evidence, and adoption analytics into a structured presentation. Part of the Communication & Content Production pillar.
This is a content assembly and generation skill. It produces a data-rich deck that the CSM customises with narrative, strategic recommendations, and relationship-aware framing. The agent builds the evidence; the human tells the story.
When to Run
- Triggered: By lo-qbr-orchestrator at T-7 before a scheduled QBR
- On-demand: When a CSM requests a QBR deck for a specific account
Core Execution Logic
Step 1: Pull Data from All Sources
Gather the complete data picture for the account:
| Data Category | Source Skill | Specific Metrics Pulled | |-------------|-------------|----------------------| | Usage and adoption | bi-usage-monitor, pa-adoption-tracker | DAU/MAU, feature adoption breadth, key workflow completions, trend direction, period-over-period change | | Health | bi-health-score | Composite score, component breakdown, trajectory over last 2+ quarters | | Value delivered | pa-value-reporter | Headline metric, supporting evidence, time-to-value benchmarks, customer-defined KPI progress | | Benchmarks | pa-benchmark-engine | Account position vs. segment peers across key metrics | | Support | Support platform | Ticket volume, resolution time, CSAT, open issues, escalation history | | Milestones | lo-milestone-tracker | Success plan progress, milestones completed, milestones overdue | | Engagement | ra-engagement-tracker | Touchpoint history, stakeholder engagement health | | Contract | CRM | ARR, renewal date, product tier, contract terms | | Feature requests | ic-feature-request-tracker | Open requests with roadmap status, resolved requests since last QBR |
Data freshness check: All data must be within 7 days of the QBR date. Flag any data source that is stale. A QBR deck with outdated data undermines credibility.
Step 2: Select Deck Template by Segment
| Segment | Slide Count | Structure | Audience Expectation | |---------|------------|-----------|---------------------| | Enterprise | 10-15 slides | Formal structure with executive summary, detailed data slides, strategic recommendations, and discussion time | Decision-makers expect data depth, benchmarking, and forward-looking strategy | | Mid-Market | 8-10 slides | Balanced structure with usage review, value highlights, support summary, and looking ahead | Operators expect actionable insights and clear next steps | | SMB | 5-6 slides (or email format) | Condensed with usage snapshot, value highlights, tips, and action items | Busy leaders expect brevity and immediate takeaways |
Step 3: Populate Slides by Template
Enterprise QBR deck:
| Slide | Content | Data Source | CSM Input Required | |-------|---------|------------|-------------------| | 1. Title and agenda | Meeting details, attendees, agenda items | Calendar, CRM | CSM may adjust agenda based on conversation goals | | 2. Executive summary | 3-5 bullet headline: health trajectory, value headline, key milestone, primary discussion topic | bi-health-score, pa-value-reporter | CSM writes the "so what" framing | | 3. Relationship overview | Key events since last QBR, milestones achieved, stakeholder changes | lo-milestone-tracker, ra-stakeholder-mapper | CSM adds context on relationship dynamics | | 4-5. Usage and adoption | Usage trend charts, feature adoption breadth, peer comparison | bi-usage-monitor, pa-adoption-tracker, pa-benchmark-engine | CSM highlights what matters to this customer | | 6. Value delivered | Headline metric, supporting evidence, ROI calculation | pa-value-reporter | CSM validates methodology and framing | | 7. Support summary | Ticket volume, resolution time, CSAT, open issues | Support platform | CSM adds context on any sensitive support history | | 8. Product roadmap | Relevant upcoming features, feature request status updates | ic-feature-request-tracker, product roadmap | CSM selects which roadmap items to highlight | | 9. Strategic recommendations | PLACEHOLDER -- CSM writes this slide | N/A | This is the most valuable slide. The agent cannot write it because it requires judgment about the account's future direction | | 10. Discussion | Open floor for customer priorities and questions | N/A | CSM prepares discussion questions based on relationship context | | 11. Action items | Captured during the meeting (populated post-QBR) | N/A | CSM captures live |
Mid-Market QBR deck follows the same logic with slides 3 and 8 consolidated and lighter data depth.
SMB QBR compresses to: usage snapshot, value highlights, 2-3 tips/recommendations, and action items.
Step 4: Generate Comparison Views
For every data slide, produce comparison context:
| Comparison Type | What It Shows | Where Used | |----------------|-------------|-----------| | Period-over-period | This quarter vs. last quarter | Every data slide -- shows trajectory | | Benchmark | Account vs. segment peers (from pa-benchmark-engine) | Usage and adoption slides -- shows relative position | | Plan vs. actual | Success plan milestones: target vs. actual dates | Milestone slide -- shows accountability | | Year-over-year | Same quarter last year (if data available) | Value slide -- shows long-term trend |
Step 5: Apply Quality Gates
Before delivering the deck:
| Gate | Check | Fail Action | |------|-------|------------| | Data freshness | All data within 7 days of QBR date | Flag stale sources. Pull fresh data if possible | | Data accuracy | Spot-check 3 key numbers against CRM source | If discrepancy found, investigate before delivering | | Customer-safe | No internal terminology (risk scores, internal notes, CSM assessments) visible | Strip all internal-only data. This is a customer-facing deck | | Completeness | All template slides populated (except CSM-input slides) | Flag any slide with missing data. Provide the slide with a "data unavailable" note rather than omitting it | | Readability | Charts are labelled, numbers are formatted, slide titles are descriptive (not just "Usage") | Reformat any chart that requires explanation to understand | | Strategic slides | CSM-input slides are clearly marked as placeholders | The agent should never fill in the strategic recommendation slide with generic advice |
Output Format
{
"account_id": "string",
"qbr_date": "2026-04-15",
"deck_format": "enterprise",
"slides_generated": 11,
"slides_requiring_csm_input": 3,
"data_freshness": {
"all_current": true,
"stale_sources": [],
"oldest_data_point": "2026-04-09"
},
"content_summary": {
"health": { "score": 74, "trend": "improving", "vs_last_qbr": "+6" },
"usage": { "dau_mau": 0.42, "feature_breadth": 0.65, "trend": "growing" },
"value": { "headline": "340 hours saved this quarter", "vs_prior_quarter": "+22%" },
"support": { "tickets": 12, "avg_resolution_hours": 4.2, "csat": 4.6 },
"benchmark_position": "Above average across key metrics. Top quartile in workflow completions"
},
"csm_action_required": [
"Write strategic recommendations (slide 9)",
"Review and customise executive summary framing (slide 2)",
"Prepare discussion questions based on relationship context (slide 10)"
],
"delivery_status": "draft_ready_for_csm_review"
}
Handoff to Human
| Decision Point | Context Provided | What the Human Decides | |---------------|-----------------|----------------------| | Strategic recommendations | Full data package, trend context, adoption gaps | What to recommend for the next quarter. This is the slide that demonstrates the CSM's value as a strategic partner, not just a data presenter | | Narrative framing | Data slides with comparison views | How to tell the story. The same data can be framed as "strong performance" or "room for improvement" depending on what the CSM wants the customer to focus on | | Slide selection | Full deck template | Which slides to keep, which to cut, which to reorder. Not every QBR needs every slide. A 45-minute QBR with 15 slides means 3 minutes per slide -- too fast for depth | | Data emphasis | All metrics with peer comparisons | Which metrics to highlight. Lead with strengths to build confidence, then address gaps constructively | | Sensitivity review | Customer-safe check applied | Whether any data point, while factually accurate, should be excluded because of customer sensitivity (e.g., a metric that declined due to a customer-side error) |
Confidence and Limitations
- High confidence for data population -- pulling metrics and formatting into slides is deterministic
- High confidence for period-over-period computation -- mathematical comparison with defined windows
- Medium confidence for chart selection and emphasis -- the skill highlights statistically significant changes, but the CSM may want different emphasis based on conversation strategy
- Medium confidence for benchmark context -- peer comparison is factual, but whether a specific benchmark motivates or discourages this customer depends on their competitive mindset
- Low confidence for narrative framing -- the skill generates factual summaries, not persuasive narratives. The CSM adds the story. A QBR deck without narrative is a data dump, not a business review
- Cannot generate the strategic recommendation slides. These require judgment about the account's future direction, relationship dynamics, and what the customer is ready to hear. This is the CSM's highest-value contribution to the QBR
- Cannot determine which slides the customer will care about most. An executive may want to skip straight to the financial summary; a technical lead may want to deep-dive into usage. The CSM adjusts the presentation flow based on the room
Dependencies
Required:
- bi-health-score (health data for trajectory slides)
- bi-usage-monitor (usage data for adoption slides)
- pa-value-reporter (value evidence for ROI slides)
- CRM API (contract data, account context)
Strongly recommended:
- pa-adoption-tracker (feature-level adoption for depth)
- pa-benchmark-engine (peer comparison data)
- lo-milestone-tracker (success plan progress)
- Support platform (support summary data)
- ic-feature-request-tracker (roadmap update status)
- ra-engagement-tracker (engagement context for relationship slides)
Downstream consumers:
- CSM (deck review and customisation -- the most important consumer. The deck is a tool for the CSM, not the final product)
- lo-qbr-orchestrator (deck delivery as part of the QBR lifecycle workflow)
- CRM (QBR activity record with deck link)
- pa-value-reporter (QBR is a validation point for value metrics -- customer reaction feeds back)
References
references/deck-guidelines.md-- Slide design principles, data visualisation standards, and customer-safe content rules
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: stephenrogan
- Source: stephenrogan/augmented-csm
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.