chaussky/ainalyst/change-strategy

SKILL: 6.4 — Define Change Strategy

BABOK chapter: 6 — Strategy Analysis Task: 6.4 Define Change Strategy MCP server: skills/change_strategy_mcp.py

What this task is about

6.4 is the culmination of Chapter 6. It synthesizes everything done in 6.1–6.3:

• From 6.1: we know the current state and business needs (BN-xxx)
• From 6.2: we know the future state and business goals (BG-xxx), and the gap analysis

is auto-imported as context — the platform cannot judge how big a gap is, so you still set each capability's gap_severity yourself in define_solution_scope. Name the element there as gap_source = "6.2:technology" (the 6.2 element this capability covers), and the platform reports which analysed gaps no capability addresses. A bare "6.2:gap_analysis" or "manual" names no element: coverage is then reported as uncheckable rather than as uncovered. Two elements, business_needs and external, are context rather than capability targets and are reported separately, excluded from the coverage count by default — unless a capability's gap_source names one of them explicitly, which pulls it back in.

• From 6.3: we know the risks (RK-xxx) and the recommendation

The task: build a substantiated transition strategy — what we're doing, how, and in what order.

Key outputs:

• {project}_change_strategy.json — machine-readable contract for 7.x, 8.x
• Markdown report — strategic document for the sponsor

When to read the references

Pipeline — 7 steps

scope_change_strategy
      ↓
define_solution_scope
      ↓
assess_enterprise_readiness
      ↓
add_strategy_option × N    ← minimum 2: one real option + do_nothing (auto-added)
      ↓
compare_strategy_options
      ↓
define_transition_states × N phases
      ↓
save_change_strategy

Step 1 — scope_change_strategy

What it does: Initializes 6.4 + auto-imports context from 6.1, 6.2 (goals and the gap analysis), 6.3.

Parameters:

• project_id — the same one used in 6.1/6.2/6.3
• change_type — transformation / process_improvement / technology_implementation / regulatory_compliance / other

<br>⚠️ 7.5 set_change_strategy asks the same question with a shorter, different set (technology / process / organizational / hybrid) and the two share no value. You will not normally meet this — 7.5 reads THIS strategy directly and refuses its own stand-in when a 6.4 strategy exists. The mapping, for the case you do, is in the 7.5 methodology. Merging the two vocabularies requires migrating strategies already on disk and is scheduled for the CLI port.

• time_horizon_months — target horizon in months
• methodology — agile / waterfall / hybrid
• source_project_ids — JSON list of project_id values from 6.1/6.2/6.3 (for auto-import)

What it returns: A summary of the imported context (BN, BG, RK, 6.2 gap elements)

• confirmation of initialization.

Automatically adds OPT-000 (do_nothing) to the list of options.

Questions to ask the BA before calling:

1. What type of change is this — a deep transformation or a targeted improvement?

2. Is there a hard deadline (regulatory, contractual)?

3. Agile, waterfall, or a hybrid approach in the organization?

Step 2 — define_solution_scope

What it does: Builds the list of capabilities with gap_severity and the explicit scope exclusions.

Parameters:

• project_id
• capabilities_json — JSON array of capability objects (see format below)
• explicitly_excluded — JSON list of what is explicitly NOT included
• scope_summary — 2–3 sentences: what we're doing and what we're not

Capability format:

{
  "name": "CRM system (base module)",
  "category": "technology",
  "description": "Replacement of the current customer management system",
  "gap_severity": "high",
  "gap_source": "6.2:technology",
  "in_scope": true
}

Categories: process / technology / data / people / org_structure / knowledge / location

Questions to ask the BA:

1. What exactly should the organization be able to do after the project that it can't do now?

2. What exactly is NOT in scope? (important to capture to prevent scope creep)

3. Which 6.2 element does this capability close? (then gap_source = "6.2:<element>",

e.g. "6.2:technology" — the eight valid elements are listed in

references/change_strategy_guide.md §4). If none, "manual".

Step 3 — assess_enterprise_readiness

What it does: Assesses 6 dimensions of organizational readiness on a 1–5 scale. Computes the final readiness_score and verdict (ready / proceed_with_caution / not_ready).

Parameters:

• project_id
• leadership_commitment — 1–5 + leadership_rationale
• cultural_readiness — 1–5 + cultural_rationale
• resource_availability — 1–5 + resource_rationale
• operational_readiness — 1–5 + operational_rationale
• technical_readiness — 1–5 + technical_rationale
• change_history — 1–5 + change_history_rationale

Interpretation:

• score ≥ 4.0 → ready
• 2.5–3.9 → proceed_with_caution (preparatory measures needed)
• < 2.5 → not_ready (readiness program first)

If the BA doesn't know how to assess a dimension: read references/readiness_guide.md §6 for questions.

Step 4 — add_strategy_option (repeat N times)

What it does: Adds a strategy option to the options registry.

OPT-000 (do_nothing) is added automatically in step 1. Add the real options (min 1).

Parameters:

• project_id
• name — option name
• strategy_type — big_bang / phased / pilot_first (not do_nothing — it already exists)
• investment_level — high / medium / low
• timeline_months — implementation timeline
• linked_risks — JSON list of RK-xxx that the option mitigates or worsens
• risk_impact — mitigates / exacerbates / neutral (for each option)
• pros — JSON list of advantages
• cons — JSON list of disadvantages

How many options are enough? At least 2 real options + do_nothing. 3–4 is optimal.

Step 5 — compare_strategy_options

What it does: Weighted comparison matrix → winner (recommended option) + opportunity cost for the rejected ones + narrative.

Parameters:

• project_id
• scores_json — JSON scoring matrix: {"OPT-001": {"alignment_to_goals": 4, "risk_mitigation": 3, ...}}
• weights_json — optionally override the default weights (must sum to 100)
• custom_criteria_json — optional additional criteria
• opportunity_cost — text: what we give up by choosing the winner instead of the rest

Default criteria and weights:

Your job (Claude): After the call — write a narrative explaining why the winner won, referencing concrete data (risks, goals, readiness).

Step 6 — define_transition_states (repeat for each phase)

What it does: Describes a transition phase — what's delivered, what's closed, what remains.

Parameters:

• project_id
• phase_number — phase number (1, 2, 3...)
• phase_name — phase name
• duration_months — duration
• capabilities_delivered — JSON list of capabilities delivered in this phase
• gaps_closed — JSON list of gap names closed after this phase
• risks_remaining — JSON list of RK-xxx risks that remain after this phase
• value_realizable — description of the value realized by the end of the phase

Rule: Each phase must deliver standalone value — otherwise rethink the slicing.

Number of phases: Depends on strategy_type:

• big_bang → 1 phase
• phased → 2–5 phases
• pilot_first → usually 2–3: pilot + rollout + (optionally) scale

Step 7 — save_change_strategy

What it does:

• Saves {project}_change_strategy.json to DATA_DIR (contract for 7.x, 8.x)
• Generates a Markdown report via save_artifact()
• Optionally: registers the solution in the 5.1 repository as a solution_scope-type node

Parameters:

• project_id
• push_to_traceability — True if you're maintaining 5.1 traceability (default: False)
• traceability_project_id — if the 5.1 repository is under a different project_id

When push_to_traceability=True:

• Creates node SOL-001 of type solution_scope with status defined

(NOT solution, which is the BABOK requirement class in the 5.1 vocabulary, and NOT status approved, which is 5.5's stakeholder-approval outcome)

• Links: SOL-001 satisfies BG-xxx (for each business goal from 6.2)

After saving — tell the BA:

1. The chosen strategy and the rationale
2. The path to the JSON (for 7.x)
3. What needs to be done about the lowest readiness dimensions before starting

Quick answers to typical BA questions

"Why consider do_nothing?" BABOK requires explicitly justifying why inaction is worse. This makes the decision defensible before the board: "we considered the status quo — here's why it's unacceptable."

"How many transition states do we need?" For phased — 2–5. More than 5 phases often signals an unclear scope or excessive granularity. One transition state = big_bang.

"If readiness_score is low — do we cancel the project?" No — it's a signal. Either change strategy_type to pilot_first/phased, or add a preparatory phase 0 (organizational readiness).

"Where's the line between 6.4 and 7.1?" 6.4 defines WHAT (capabilities and scope) and WHEN (phases). 7.1 defines exactly HOW — detailed requirements for each capability.

Relationship to other tasks