LAB QUOTA · OK
[ change-request:// ] experimental
cat: ai model: @cf/meta/llama-3.1-8b-instruct

Scope change you need to formalise → get a proper Change Request doc (impact, owner, dependencies, decision needed by).

// system prompt
You write Change Request documents. User provides change details. Output:

  # Change Request: <one-line summary>
  CR-ID: CR-<sequential, leave as CR-XXX if unknown>
  Date raised: <today's date or as provided>
  Raised by: <who>
  Status: Proposed | Approved | Rejected | Deferred

  ## Change Requested
  <2-3 sentences. What's being changed, what was the original commitment.>

  ## Business Reason
  <1-3 sentences. Why this came up now. What business outcome it enables.>

  ## Impact Analysis
  ### Time impact
  <effect on schedule>
  ### Cost impact
  <effect on cost / budget>
  ### Resource impact
  <effect on team capacity>
  ### Quality impact
  <what gets de-prioritised to make room — or what risk we take on>
  ### Risk impact
  <new risks introduced; existing risks affected>

  ## Options
  1. **Approve as requested** — pros / cons
  2. **Approve with modification** — what would change, pros / cons
  3. **Defer / Reject** — pros / cons

  ## Recommendation
  <which option, with one-line reason>

  ## Decision Needed
  - Decision owner: <role>
  - Decision deadline: <date — gated on the project's next critical path event>
  - If no decision by deadline: <default action — usually "treated as Rejected">

  ## Affected Stakeholders
  • <list of people who need to know the decision>

Rules:
- The CR is a DECISION request, not a status update. Lead with the change + decision needed.
- Always offer 3 options including a "Defer / Reject" baseline.
- Recommend ONE option — don't hide behind neutrality.
- "Decision Deadline" must be anchored to a project event (not a vibe).
- Don't pad. A well-written CR is 1 page.
⚡ powered by Cloudflare Workers AI · quota deducted on success
// sample output
# Change Request: Add SAML SSO to the customer-portal refresh
CR-ID: CR-XXX
Date raised: today
Raised by: Sales VP (via Slack)
Status: Proposed

## Change Requested
Add SAML SSO support to the customer-portal refresh release scheduled for 5 Aug 2026. SSO was originally deferred to the Q4 release based on the prioritisation done in March.

## Business Reason
Two enterprise prospects in late-stage sales (combined ~$220k ARR potential) have stated SSO is a hard requirement to close. Without SSO at launch, both deals slip into Q4 or churn entirely. SSO at launch would unlock both this quarter.

## Impact Analysis

### Time impact
SAML SSO adds approximately 12-15 dev-weeks of effort (integration + testing + security review). At current 9 dev-weeks of headroom in the plan, this means either an additional 3-6 dev-weeks of work OR pulling other in-scope items.

### Cost impact
No additional licensing cost (using existing identity-provider integration patterns). Engineering cost = ~12-15 dev-weeks of capacity reallocation.

### Resource impact
No new headcount. Requires re-prioritisation of the QoL bundle (10 dev-weeks) and possibly partial deferral of the notification center (10 dev-weeks).

### Quality impact
If approved as requested, the QoL bundle moves to Q4. Notification center stays in scope but with reduced polish budget (~2 dev-weeks of polish dropped).

### Risk impact
New risks:
• SAML SSO integration touches authentication — small chance of breaking existing auth flows. Mitigation: feature-flag the SSO path, opt-in for the two prospect customers only at launch.
• Compressed timeline = compressed security review window. May force a tradeoff.

Existing risks affected:
• Launch gate (currently green) moves to amber given new scope at T-12 weeks.

## Options

1. **Approve as requested** — Add SAML SSO; defer QoL bundle to Q4.
   Pros: Unlocks $220k ARR potential. Sales VP gets cover for both deals.
   Cons: QoL items affect every customer; the SSO benefit accrues to two specific prospects.

2. **Approve with modification** — Build SSO behind a feature flag with limited rollout (the 2 prospect customers only). Keep QoL items. Sales VP confirms the two deals close on Limited-Pilot status, with general availability in Q4.
   Pros: Both wins — unblocks the deals + retains the QoL items. Risk-managed via feature flag.
   Cons: Slightly more complex coordination with the 2 prospect customers. Sales VP needs to confirm Limited-Pilot is acceptable to them.

3. **Defer / Reject** — Keep current scope. SSO remains Q4. Two deals slip or churn.
   Pros: Preserves the existing plan and the QoL items.
   Cons: $220k ARR walks. Sales VP relationship.

## Recommendation
**Option 2 (Approve with modification).** Captures the revenue opportunity without sacrificing the broader customer benefit of the QoL items. Limited-Pilot framing also reduces security-review pressure since the SSO surface area at launch is bounded.

## Decision Needed
- Decision owner: VP Product (in consultation with Sales VP + VP Engineering)
- Decision deadline: End of this week. Beyond that, integration work must start to land in W12 — every day of delay reduces the QoL items proportionally.
- If no decision by deadline: treated as Rejected (Option 3); we proceed with the existing plan and SSO stays Q4.

## Affected Stakeholders
• VP Product (decider)
• VP Engineering (technical owner)
• Sales VP (originator + confirmer)
• Lead engineer for the auth workstream
• Compliance lead (security review impact)
• The two prospect customers (must know if Limited-Pilot is on offer)
// powered by cloudflare workers ai · quota deducted on success ← back to catalog