Rob Dull Tools AI ProdOps Document Flow
Agentic PDLC • Real Toolchain, Real Handoffs

The document flow, tool by tool.

What each of the nine tools in the chain produces, and where it goes next: every handoff below is a real same-origin localStorage key written by one tool and read by the next, not a conceptual diagram. Each step's primary output is the document that gates the next step; secondary outputs are the other real artifacts produced alongside it. Four primary outputs carry the most weight for program stakeholders: the BRD, the Roadmap, the Feature list (the extended backlog), and the Jira export. Every step links to a real sample run: click through to see real output, not a mockup.

↗ See the full aspirational system in the Network Diagram · New here? Start with the plain-language explainer ↗
View:
Primary and secondary inputs/outputs for the nine tools: wired and real.
Toolchain: structured state produced by another working tool
Fixture evidence: fictional PetHealth records shaped like potential enterprise evidence
Human-supplied: context entered by the practitioner
Target architecture: proposed data or service connection that is not implemented
Deterministic check: a specific condition recalculated in code, not asserted by the model
The four primary outputs that matter most
STEP 07
BRD
Standard-template Business Requirements Document, one per funded epic.
STEP 06
Roadmap
Release windows across up to 18 months, funded epics first.
STEP 08
Feature list
MoSCoW'd features + ordered stories, the extended backlog.
STEP 09
Jira export
Bulk-import CSV: priority, points, sprint, dependency links.
01
A rough user description becomes a detailed enterprise persona: JTBD statement, org context, key interactions, relationship friction.
Inputs
Primary
MCP
Research repo (Dovetail/Notion) Analytics Customer support (Zendesk)
Outputs
Primary
Persona

Primary risk: A plausible composite can be mistaken for research.

Required review: Confirm which attributes are supported, inferred, or invented.

External inputs Sales transcripts (Gong) Stakeholder input (seen in the network diagram, not wired into this tool)
↗ Open tool No dedicated sample-run page: the tool itself is the demo.
02
Stage-by-stage canvas (actions, goals, touchpoints, emotions, and pain points) seeded from a persona.
Inputs
Primary
Toolchain
Persona
Secondary
MCP
Research repo (Dovetail/Notion) Customer support (Zendesk)
Outputs
Primary
Opportunity seeds
Secondary
Journey map (JSON / Copy for AI)

Primary risk: Team assumptions can be rendered with the visual authority of research.

Required review: Separate observed evidence from hypotheses.

External inputs Usage analytics NPS / CSAT (seen in the network diagram, not wired into this tool)
03
Initiative Intake Portfolio Tradeoff
Frames one Initiative against competitors, then decomposes it into Business and Architecture Epics with Demand/Feasibility/Viability assumptions and a benefits-realization plan.
Inputs
Primary
Toolchain
Opportunity seeds
Secondary
MCP
Internal comms Financial tracking External comms (NPS, reviews) Analytics
Recommended
Strategic KPIs
Outputs
Primary
Epics (business + architecture)
Secondary
D/F/V assumption map Benefits-realization plan

Primary risk: Missing portfolio context may be filled with plausible competitors, budgets, or benefits.

Required review: Validate the comparison set, epic boundaries, assumptions, KPIs, baselines, and proposed instrumentation.

External inputs Product backlog (dedup) Win/loss data Competitive signals Org strategy / OKRs (seen in the network diagram, not wired into this tool)
04
Business Cases Justify
One case per epic: value, cost, strategic alignment, and a rough order-of-magnitude effort estimate that anchors the Epic Prioritizer's job size.
Inputs
Primary
Toolchain
Epics (business + architecture)
Secondary
MCP
Financial tracking External comms
Outputs
Primary
Business Case per epic

Primary risk: Generated cost, benefit, and effort statements can appear more certain than the evidence supports.

Required review: Replace unsupported estimates and financial precision with sourced values or explicit unknowns.

External inputs Org strategy / OKRs Market research Win/loss data (seen in the network diagram, not wired into this tool)
05
Reads each epic's Business Case as its candidate record: value statement, cost/benefit, strategic alignment, and the rough effort (LOE) from Step 04, and scores every one on WSJF and RICE components in one pass. Editable weights recompute the board live: raising risk-reduction can flip an architecture epic to #1 with no re-run.
Inputs
Primary
Toolchain
Business Cases (candidate records)
Secondary
MCP
Internal comms (leadership weighting) Customer support (reach/severity)
Outputs
Primary
Epic Ranking + funding line
Secondary
RICE alternative scoring

Primary risk: Correct arithmetic can lend false authority to weak input values.

Required review: Own every input, weight, and funding decision.

External inputs NPS / CSAT Competitive signals ITSM / incidents (seen in the network diagram, not wired into this tool)
06
Sequences every prioritized epic (funded first, the rest as capacity frees) into release windows across up to 18 months. Dependencies stay light: named approvals or genuine architecture-epic runway only.
Inputs
Primary
Toolchain
Epic Ranking + funding line
Secondary
MCP
Delivery systems (sprint velocity) Internal comms (approvals)
Outputs
Primary
Roadmap

Primary risk: Release windows can imply capacity knowledge the workflow does not possess.

Required review: Confirm capacity, runway, dependencies, commitments, and rollout assumptions.

External inputs Delivery metrics (DORA) Team health surveys
External outputs Stakeholder comms (roadmap narrative) (seen in the network diagram, not wired into this tool)
07
Standard-template BRD for one funded epic: requirements core (scope, prioritized business requirements, high-level NFRs) plus the executive package (summary, objectives, stakeholders, cost-benefit, sign-off). Refuses epics below the funding line.
Inputs
Primary
Toolchain
Roadmap
Secondary
MCP
Internal comms (stakeholders, sign-off) Analytics (what's instrumentable today)
Outputs
Primary
BRD

Primary risk: NFR thresholds, costs, stakeholders, and sign-offs may be inferred or invented.

Required review: Confirm scope, requirement quality, evidence, assumptions, NFRs, financials, stakeholders, and approval authority.

External inputs Regulatory signals ITSM / incidents
External outputs Regulatory impact assessment (seen in the network diagram, not wired into this tool)
08
Turns the BRD into the full delivery-doc suite: MoSCoW'd Features with feature-level NFRs and refined sizing, ordered Stories with BDD acceptance criteria and dependsOn sequencing, a solution outline, security/UAT plan, and dev/user docs. Requirement coverage is computed, not generated.
Inputs
Primary
Toolchain
BRD
Secondary
MCP
Customer support (edge cases → acceptance criteria)
Organizational
DoR / DoD Style guide Current-state architecture
Outputs
Primary
Feature list (extended backlog)
Secondary
Security & UAT plan Developer & user docs

Primary risk: Identifier coverage can be mistaken for adequate design or requirement satisfaction.

Required review: Engineers, architecture, security, QA, and product stakeholders must assess substance.

External inputs Regulatory signals ITSM / incidents (seen in the network diagram, not wired into this tool)
09
Sprint Planner Deliver
MVP scope, a sprint-by-sprint allocation, and a RAID log. Capacity AND dependsOn ordering are both recomputed in code: a story landing no earlier than its dependency raises a visible warning wherever the plan is consumed.
Inputs
Primary
Toolchain
Feature list (extended backlog)
Also carries the DoR/DoD flag through from Step 08. No dedicated MCP evidence yet.
Outputs
Primary
Jira bulk-import CSV
Terminal: exports to Jira, no further handoff.
Secondary
Sprint Plan RAID log

Primary risk: The plan can only check the estimates and dependencies it was given. It cannot discover missing work or hidden capacity constraints.

Required review: Delivery leadership owns the plan and must inspect the CSV before import.

External inputs Delivery metrics (DORA / velocity) Team retros Team health surveys
External outputs Release notes ROI telemetry monitor (30/60/90-day) (seen in the network diagram, not wired into this tool)
Off-chain by design, not part of the numbered flow
Strategic alignment: the layer the whole portfolio should trace back to, not a step any one initiative passes through. Cascades a V2MOM into an X-matrix.
Measurement: watches whether the ops system itself is healthy across every initiative, not any single one. In development.
★ = primary output. "Primary/Secondary" labels rank each step's own inputs by importance, independent of chip color below
Teal: a same-origin tool handoff (this step's output, or an upstream step's, consumed directly)
Gold: MCP evidence source, machine-fetched, fixture-based
Purple: organizational input, human-supplied, never generated
Dashed orange: recommended for this flow, not yet built
Dotted steel: external/tertiary, seen only in the network diagram, not wired into this flow
Deterministic check: a specific condition recalculated in code, not asserted by the model
Link to a real sample run, not a mockup
What's real vs. recommended: every chip above except "Strategic KPIs" and the dotted external chips reflects code that runs today. The MCP evidence is fixture-based (PetHealth-flavored, keyless demo data with the same shape a live connector would have) but genuinely wired into each tool's prompts, and every handoff key is a real localStorage write you can inspect in devtools after a run. Every sample-run link goes to that tool's own built-in demo: no API key, real output, not staged for this page.
What "External (Network)" adds: toggling to that view surfaces the data sources and outputs that exist in the full Network Diagram (CRM/win-loss, competitive intel, regulatory signals, ITSM, delivery telemetry, retros, and their downstream outputs: release notes, stakeholder comms, ROI telemetry) but are not wired into any of the nine tools on this page. They're the natural next inputs/outputs, not fiction: each one corresponds to a real node in the network diagram with its own MCP recommendation.