FinOps Dashboard `runbooks finops dashboard`¶

🧭 Purpose

The runbooks quickstart for the runbooks finops dashboard command with --mode executive|architect|sre for 🧑‍💼 Executive, 🏗️ Architect, 🔧 SRE

1. Quick Start: Optimal Defaults (Persona-Driven) 🚀¶

Executive Mode (C-Suite Focus)

Override defaults as needed: --timeframe daily (not monthly) --top-n 10 (not 5) --cost-threshold 5000 (not 0) --validation-level basic (not mcp)

runbooks finops dashboard --profile $AWS_PROFILE --mode executive --timeframe daily --top-n 10 --validation-level basic --export csv --export pdf

Auto-applies: daily, top-5, threshold-5000, mcp validation

# runbooks finops dashboard --profile $AWS_PROFILE --mode executive --export csv

Architect Mode (Infrastructure Teams)

Infrastructure Analysis: --top-n 20 (not 10) --timeframe monthly (trend analysis) --output-format both (table + tree) --activity-analysis (decommission signals)

runbooks finops dashboard --profile $AWS_PROFILE \
  --mode architect --activity-analysis \
  --filter 'service:EC2,RDS,ELB' --top-n 20 --output-format tree \
  --export csv --export markdown

Auto-Applied Defaults: top-10, monthly, both (table+tree)

# runbooks finops dashboard --mode architect --activity-analysis --export csv

SRE Mode (Operations)

Full operational context & cost spike: --timeframe weekly (not monthly) --top-n 10 (not 5) --sort-by change (not current - Anomalies first) --validation-level mcp (not basic) --verbose (debug context) --export json --export csv (programmatic alerting)

runbooks finops dashboard --mode sre --timeframe weekly \
  --filter 'cost>100' --validation-level mcp \
  --top-n 10 --sort-by change --verbose \
  --export json --export csv

Auto-applied Defaults: weekly, sort-by-change, mcp validation

# runbooks finops dashboard --mode sre --filter 'cost>1000' --export json

ℹ️ Note: Replace profile names and filters according to your environment and tagging standards.

Multi-Account (All Modes) compliance audit - validate cost allocation across org

Executive: Multi-account compliance audit

runbooks finops dashboard --all-profile management --mode executive --validation-level mcp

Architect: Organization-wide infrastructure review

runbooks finops dashboard --all-profile management --mode architect --activity-analysis --validation-level basic

SRE: Multi-account cost spike detection

runbooks finops dashboard --all-profile management --mode sre --filter 'cost>1000'

Verify Smart Defaults Applied

Show which defaults were auto-applied (use --verbose)

runbooks finops dashboard --mode executive --verbose 2>&1 | grep "Applied"

## Expected output:
# Applied executive persona default: timeframe=daily
# Applied executive persona default: top_n=5
# Applied executive persona default: cost_threshold=5000
# Applied executive persona default: validation_level=mcp

Override Defaults (User Control Preserved): Executive mode but with architect-style parameters: monthly instead of daily, top-20 instead of top-5, cost-threshold 0 instead of 5000, basic instead of mcp

runbooks finops dashboard \
  --mode executive \
  --timeframe monthly \ 
  --top-n 20 \ 
  --cost-threshold 0 \ 
  --validation-level basic

2. Cross-Persona Command Overview (1 CLI & 3 modes) 🧰¶

Category	Option(s)	🧑‍💼 Executive	🏗️ Architect	🔧 SRE / CloudOps
Scope selector	`--profile` / `--all-profile`	Uses billing/central profile	Uses platform / env profiles	Uses app / ops / env profiles
	`--timeframe`	`monthly`, `quarterly`	`weekly`, `monthly`, `quarterly`	`daily`, `weekly`
	`--cost-metric`	`amortized` / `blended`	`unblended` / `amortized`	`unblended`
	`--accounts`, `--services`	Typically not used	Heavily used	Heavily used
Filter / focus	`--filter`	High-level (e.g. `cost>1000`)	By service/pattern	By workload/resource
	`--cost-threshold`, `--top-n`, `--sort-by`	Top-N spenders	Top patterns / components	Top anomalies / spikes
Mode & analysis	`--mode`	`executive`	`architect`	`sre`
	`--activity-analysis`	Optional (for summary)	Often enabled	Often enabled
	`--validation-level`	`mcp` or `strict` for board packs	`mcp` for platform reports	`basic` or `mcp` for ops
Output	`--export`	`pdf`, `markdown`, sometimes `csv`	`csv`, `markdown`, `json`	`csv`, `json`, `markdown`
	`--summary-mode`, `--output-format`	`table` or `both`	`both`	`table` or `tree`
	`--screenshot`	For snapshot in Confluence	Sometimes for architecture reviews	Sometimes for incident post-mortems

🧩.1 Essential Options – Core Workflow Parameters

Option	Role	Notes
`--profile TEXT`	Scope selector	AWS profile; auto-detects Organizations where applicable.
`--all-profile TEXT`	Org-wide scope	Management profile for multi-account Organizations: MANAGEMENT_PROFILE, BILLING_PROFILE, CENTRALISED_OPS_PROFILE
`--timeframe daily\\|weekly\\|monthly\\|quarterly`	Time window	Analysis period (default: `monthly`).
`--cost-metric blended\\|unblended\\|amortized`	Cost metric	Calculation method as defined in your billing source.
`--filter TEXT`	Targeting	Filter DSL (repeatable).
`--validation-level basic\\|mcp\\|strict`	Data quality	Validation level (`mcp`: ≥99.5% per help text).
`--export csv\\|json\\|pdf\\|markdown`	artifacts	Export format (repeatable).
`-v, --verbose`	Logging	Diagnostic logging.

🧩.2 Advanced Options – Power User Controls

Option	Role	Notes
`--services TEXT`	Service subset	Comma-separated list of AWS services.
`--accounts TEXT`	Account subset	Comma-separated list of AWS account identifiers.
`--cost-threshold FLOAT`	Cost floor	Minimum cost filter.
`--top-n INT`	Top-N control	Range 1–50 (default: 10).
`--sort-by current\\|previous\\|change`	Ordering	Sort order for service tables.
`--show-zero-cost`	Visibility	Include services with 0 cost.
`--show-empty`	Visibility	Include resource types with 0 resources.
`--summary-mode table\\|tree\\|both\\|none`	Summary style	Select source of summary.
`--output-format both\\|table\\|tree`	Visualisation	Choose output representation.
`--screenshot`	Snapshot	Playwright screenshot (requires HTML output).
`--dry-run`	Safeguard	Executes in dry-run mode (default: true).

🧩.3 Feature Flags – Experimental

Flag	Role	Notes
`--activity-analysis`	Activity signals	Enables E1–E7, R1–7, S1–7 signals as defined by the tool.
`--mode executive\\|architect\\|sre`	Persona lens	Controls dashboard presentation style.

3. Runbook: Executive Mode (`--mode executive`) 🧑‍💼¶

🎯 Executive monthly review runbook (C-suite)

3.1 Typical scenarios¶

Monthly / quarterly spend review.
Identify services or accounts above a cost threshold.
Export clean tables for board packs / Confluence.

3.2 Step-by-step: CLI workflow¶

Step 1 – Choose scope & timeframe

Use your billing or central profile and a business-relevant period:

runbooks finops dashboard \
  --mode executive \
  --profile $BILLING_PROFILE \
  --timeframe monthly \
  --cost-metric amortized

Key points (directly from your options):

--profile $BILLING_PROFILE Uses the profile that can see consolidated billing (as documented in your help text).
--timeframe monthly Uses a business-aligned period; matches monthly financial cycles.
--cost-metric amortized Shows cost including amortized commitments so executives see “true” cost per period.

Step 2 – Focus on material spend

Apply a threshold and top-N view:

runbooks finops dashboard \
  --mode executive \
  --profile $BILLING_PROFILE \
  --timeframe monthly \
  --cost-metric amortized \
  --cost-threshold 1000 \
  --top-n 20 \
  --sort-by current

You’re using only documented flags:

--cost-threshold 1000 → only services/accounts above 5,000 (currency same as billing).
--top-n 20 → short, manageable list.
--sort-by current → sorts by current period cost.

Step 3 – Validate & generate export

Executives need trusted numbers and offline artifacts:

runbooks finops dashboard \
  --mode executive \
  --profile $BILLING_PROFILE \
  --timeframe monthly \
  --cost-metric amortized \
  --cost-threshold 1000 \
  --top-n 20 \
  --sort-by current \
  --validation-level mcp \
  --summary-mode table \
  --output-format table \
  --export pdf \
  --export markdown

--validation-level mcp Uses your MCP validation level (per help: “mcp: ≥99.5%”) for exec-grade reliability.
--summary-mode table, --output-format table Forces a clearly structured tabular summary aligned with how finance works.
--export pdf, --export markdown Creates a PDF and Markdown snapshot appropriate for:
PDF → board packs / email attachment.
Markdown → Confluence / internal docs.

You can run with --dry-run first (default is true) to verify logic, then re-run without --dry-run if the implementation treats it as “no side effects”.

Step 4 – Optional: screenshot for visual dashboard

If the tool renders HTML dashboards, you can add:

--screenshot

…to capture a viewport image via Playwright (per your help).

3.3 Expected outputs (Exec mode)¶

Using only what your CLI advertises, output will include:

Summarised tables (per --summary-mode table):
Top services / accounts by cost above --cost-threshold.
Sorted as per --sort-by (current, previous, or change).
Exports:
*.pdf with the executive-view dashboard (C-suite phrasing, per --mode executive description).
*.md with tables that can be pasted into Confluence or emails.

3.4 Executive acceptance checklist ✅¶

For each monthly run, confirm:

✅ Used billing or central profile (--profile $BILLING_PROFILE or similar).
✅ --timeframe matches the reporting period (e.g. monthly).
✅ --cost-metric chosen consistently (e.g. amortized).
✅ --validation-level is mcp or strict for executive packs.
✅ --cost-threshold and --top-n keep the view focused on meaningful items.
✅ --export includes at least one shareable format (pdf, markdown).

4. Runbook: Architect Mode (`--mode architect`) 🏗️¶

🏗️ Architect / Head of Platform runbook

4.1 Typical scenarios¶

Analyze compute + data tier together (per your Quick Start example).
Understand service mix and cost concentration at platform level.
Prepare data for architecture reviews and platform backlogs.

4.2 Step-by-step: infra dependency analysis¶

Step 1 – Platform scope & lens

runbooks finops dashboard \
  --mode architect \
  --profile $CENTRALISED_OPS_PROFILE \
  --timeframe monthly \
  --cost-metric blended \
  --services EC2,RDS,ELB

Here you’re:

Using --mode architect → infrastructure analysis lens.
Using a central/platform profile (based on your help text “$CENTRALISED_OPS_PROFILE”).
Narrowing to main infra services per your example: --services EC2,RDS,ELB.

Step 2 – Add activity signals & structure

runbooks finops dashboard \
  --mode architect \
  --profile $CENTRALISED_OPS_PROFILE \
  --timeframe monthly \
  --cost-metric blended \
  --services EC2,RDS,ELB \
  --activity-analysis \
  --top-n 50 \
  --sort-by change \
  --summary-mode both \
  --output-format both

All using defined options:

--activity-analysis Enables your E1–E7 / R1–R7 / S1–S7 signals (as per help text).
--top-n 50 Enough to see broader patterns.
--sort-by change Focuses on what changed, useful for architecture reviews.
--summary-mode both, --output-format both Gives tables and tree views so architects see both aggregate and structural views.

Step 3 – Export & hand off to platform backlog

Architects typically need machine-readable exports and human-readable reports:

runbooks finops dashboard \
  --mode architect \
  --profile $CENTRALISED_OPS_PROFILE \
  --timeframe monthly \
  --cost-metric blended \
  --services EC2,RDS,ELB \
  --activity-analysis \
  --top-n 50 \
  --sort-by change \
  --summary-mode both \
  --output-format both \
  --validation-level mcp \
  --export csv \
  --export markdown

--export csv → ingest into spreadsheets or data tools; attach to Jira epics or ADRs.
--export markdown → architecture review notes / Confluence.
--validation-level mcp → ensures that optimization decisions rely on validated numbers.

4.3 Expected outputs (Architect mode)¶

From the options you defined, architects get:

Service-level breakdowns (EC2/RDS/ELB here) with:
Current cost, previous period cost, and changes (per --sort-by).
Activity-analysis signals:
E1–E7 / R1–R7 / S1–S7 marks per your flag description.
Dual view:
Table + tree summarizations (--output-format both) to see:
- Aggregated stats for each service.
- Hierarchical or grouped structures relevant to infra.

4.4 Architect acceptance checklist ✅¶

For each review:

✅ Used platform / central profile (e.g. $CENTRALISED_OPS_PROFILE).
✅ --mode architect is set (infra lens).
✅ Core infra services selected using --services where needed.
✅ --activity-analysis used when analyzing behavior and patterns.
✅ --top-n and --sort-by tuned to focus on meaningful platform components.
✅ --export csv used when feeding into roadmaps / backlog tooling.

5. Runbook: SRE / CloudOps Mode (`--mode sre`) 🔧¶

🔧 SRE / CloudOps cost spike runbook

5.1 Typical scenarios¶

Investigate cost spikes in specific services, namespaces, or workloads.
Combine dashboard view with MCP validation for cross-check.
Identify candidates for rightsizing or cleanup.

5.2 Step-by-step: spike investigation¶

Step 1 – Narrow to recent spike window

runbooks finops dashboard \
  --mode sre \
  --profile OPERATIONS_PROFILE \
  --timeframe daily \
  --cost-metric unblended \
  --cost-threshold 1000 \
  --top-n 50 \
  --sort-by change

You are:

Using --mode sre → operations deep-dive presentation.
Using --timeframe daily → for recent spike analysis.
Using --cost-metric unblended → see raw, current-period charges.
Using --cost-threshold 1000 → focus on items above 1,000 in the spike window.
Using --sort-by change → highlight what changed.

(Replace OPERATIONS_PROFILE with whatever ops profile name matches your config; your help text does not prescribe a fixed name, so keep it consistent internally.)

Step 2 – Add MCP cross-validation & verbose logs

Your Quick Start example explicitly includes MCP:

runbooks finops dashboard \
  --mode sre \
  --profile OPERATIONS_PROFILE \
  --timeframe daily \
  --cost-metric unblended \
  --cost-threshold 1000 \
  --top-n 50 \
  --sort-by change \
  --validation-level mcp \
  --activity-analysis \
  --summary-mode table \
  --output-format table \
  --verbose

All options here are from your help:

--validation-level mcp → cross-checks with MCP-level validation.
--activity-analysis → attaches E1–E7 / R1–R7 / S1–S7 to each entry.
--summary-mode table, --output-format table → SRE-friendly view.
--verbose → logs diagnostic details for incidents and post-mortems.

Step 3 – Export for incident and follow-up

runbooks finops dashboard \
  --mode sre \
  --profile OPERATIONS_PROFILE \
  --timeframe daily \
  --cost-metric unblended \
  --cost-threshold 1000 \
  --top-n 50 \
  --sort-by change \
  --validation-level mcp \
  --activity-analysis \
  --summary-mode table \
  --output-format table \
  --export csv \
  --export json \
  --export markdown

This yields:

CSV → direct inspection in spreadsheets or further analysis scripts.
JSON → integration with other tooling (automation or incident systems).
Markdown → incident summary / post-mortem sections.

You can keep --dry-run if the implementation treats it as a no-side-effect analysis-only mode.

5.3 Expected outputs (SRE mode)¶

Per the --mode sre description (“Operations deep-dive”):

SRE-focused tables of resources/services that crossed your --cost-threshold.
Each row annotated with:
Current vs previous cost (given --sort-by change).
Activity signals (--activity-analysis) for operational behaviors.
Exported CSV/JSON to feed into:
Shell/automation scripts.
Jupyter notebooks for deeper investigations.

5.4 SRE acceptance checklist ✅¶

For each spike investigation:

✅ --mode sre is set (ops lens).
✅ --timeframe narrow enough (e.g. daily) to isolate the spike.
✅ --cost-threshold tuned so spike contributors are visible.
✅ --validation-level mcp used when cross-checking is required.
✅ --activity-analysis enabled for behavior insights.
✅ --export csv/json used where automation or Jupyter follow-up is expected.

6. Cross-mode quick reference matrix (CLI usage only)¶

Persona	Mandatory Flags	Common Extras	Typical Exports	Key Quality Gates
Executive 🧑‍💼	`--mode executive` `--profile $BILLING_PROFILE` `--timeframe monthly \\| quarterly` `--cost-metric amortized \\| blended`	`--cost-threshold` `--top-n` `--sort-by current \\| change` `--validation-level mcp` `strict`	`--export pdf` `--export markdown`	MCP/strict validation, stable timeframes, focused threshold & top-N
Architect 🏗️	`--mode architect` `--profile $CENTRALISED_OPS_PROFILE` `--timeframe monthly \\| quarterly`	`--services` `--activity-analysis` `--top-n` `--sort-by change` `--summary-mode both` `--output-format both`	`--export csv` `--export markdown`	Platform/profile scope, service filters, activity-analysis enabled
SRE / CloudOps 🔧	`--mode sre` `--profile OPERATIONS_PROFILE` `--timeframe daily \\| weekly` `--cost-metric unblended`	`--cost-threshold` `--top-n` `--sort-by change` `--activity-analysis` `--validation-level mcp` `--verbose`	`--export csv` `--export json` `--export markdown`	Spike-focused timeframes, explicit thresholds, detailed logs, exports for automation

FinOps Dashboard runbooks finops dashboard¶