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

Paste your team's commitments + workload → get a re-allocation plan that levels the load without slipping critical dates.

// system prompt
You re-level team workload. User provides per-person commitments + estimates. Output:

  ## Workload audit (before)
  PERSON   ASSIGNED   CAPACITY   DELTA
  ------   --------   --------   -----
  <table>

  ## Issues to address
  • Over-committed: <names with how much>
  • Under-committed: <names with how much>
  • Specialist-skill bottlenecks: <work that can't move easily>

  ## Proposed re-allocation
  Move:
  • <item> from <person> to <person> — savings: <N weeks> · feasibility: <high/medium/low — why>

  Descope or defer:
  • <item> — <reason>

  Hire / borrow:
  • <type of work> — <recommendation>

  ## Workload after re-allocation
  PERSON   ASSIGNED   CAPACITY   DELTA
  ------   --------   --------   -----
  <updated table>

  ## Items that CANNOT be moved (and why)
  • <item> — <skill / dependency reason>

  ## Watch-outs
  • <e.g. pairing increases throughput on the work but reduces total team throughput — flag>

Rules:
- Capacity is whatever the user implies (default 12 weeks for a quarter; 1 person = 1 week per week minus overhead).
- Reductions to over-committed people must come from real options, not "work harder".
- "Feasibility" of moving work is HIGH if generalist; LOW if specialist or deep context required.
- Don't suggest moving mentorship time — it's a leverage activity, not extractable.
- If even with re-allocation the math doesn't close, recommend descoping (with which items + why).
- "Hire / borrow" is OK as a recommendation but flag it as a slow lever.
⚡ powered by Cloudflare Workers AI · quota deducted on success
// sample output
## Workload audit (before)
PERSON   ASSIGNED   CAPACITY   DELTA
------   --------   --------   -----
Alice    14         12         +2  (over)
Bob       8.4       12         -3.6 (under)
Carol    11         12         -1  (roughly on plan)
Dave      7         12         -5  (under)
Eve      18         12         +6  (over)

## Issues to address
• Over-committed: Eve (+6 weeks), Alice (+2 weeks).
• Under-committed: Dave (-5 weeks), Bob (-3.6 weeks).
• Specialist-skill bottlenecks: Auth migration (Alice = primary; Dave pairing). Frontend redesign (Carol + Eve specialist work). Compliance prep (Alice — domain knowledge).

## Proposed re-allocation
Move:
• Bug burn-down (3 weeks) from Dave to Bob → savings: frees 3 weeks of Dave's pairing capacity, balances Bob's load. Feasibility: HIGH (generalist work, well-defined).
• 4 weeks of Eve's "new onboarding flow" → Dave (paired with Carol on review). → savings: 4 weeks off Eve. Feasibility: MEDIUM (Dave is junior; needs Carol review checkpoint at week 1 and 3).
• 2 weeks of Eve's remaining onboarding-flow work → Bob (he has 0.6 weeks freed + room for ~2 weeks more). Feasibility: MEDIUM (Bob has the skill; needs a context handover from Eve).
• 2 weeks of Alice's compliance review prep → Carol (Carol has 1 week of slack and the comms skill for compliance interaction). Feasibility: MEDIUM (Carol is not the domain expert; Alice needs to spend ~3 days of context transfer).

Descope or defer:
• None at this level — the math closes via re-allocation.

Hire / borrow:
• No immediate need. Consider hiring an additional mid eng for Q4 if Q3 patterns repeat — the team is running too close to the line.

## Workload after re-allocation
PERSON   ASSIGNED   CAPACITY   DELTA
------   --------   --------   -----
Alice    12         12          0  ✓
Bob      13.4       12         +1.4 (slightly over — manageable)
Carol    13         12         +1  (slightly over — manageable)
Dave     10         12         -2  (small buffer — good for unplanned work)
Eve      12         12          0  ✓

## Items that CANNOT be moved (and why)
• Auth migration (Alice → Dave is already pairing) — this is specialist work; Alice can't hand off.
• Frontend redesign (Carol + Eve) — both are specialists for the redesign; Eve's 10 weeks here is not extractable.
• Mentorship (Alice) — kept off the moveable list intentionally; it's a leverage activity.
• On-call rotation (Bob) — fixed rotation, can't reassign mid-quarter.
• Sprint planning chair (Alice) — process leadership, not movable.

## Watch-outs
• Dave now has 4 weeks of stretch work (onboarding flow). Build in a Carol checkpoint at week 1 and 3 of that work — don't leave him solo.
• Bob is +1.4 vs capacity; if his on-call rotation produces > expected incidents, he'll overflow. Have an explicit pull-the-cord plan.
• Total team is running ~96% utilised after re-allocation. Almost no buffer for surprises. Flag to leadership: this is sustainable for ONE quarter, not three.
// powered by cloudflare workers ai · quota deducted on success ← back to catalog