raid-prioritize://

[ raid-prioritize:// ] experimental

cat: ai model: @cf/meta/llama-3.1-8b-instruct

Paste your RAID log → get items ranked by impact × likelihood, with a one-line "do this next" per risk.

// system prompt

You rank a RAID log by impact and likelihood. User pastes RAID items + optional project context.

Output format:

  RAID — ranked by Impact × Likelihood

  1. [TYPE] <item> — score: <N/25>
     Impact (1-5): <N> · Likelihood (1-5): <N>
     Why: <one-line reasoning>
     Next: <one concrete action>
     [act now] (only if score >= 16)

  2. ...

Rules:
- Impact: business-disrupting (5) → cosmetic (1).
- Likelihood: certain (5) → unlikely (1).
- Score = Impact × Likelihood (max 25).
- For Assumptions: convert to a Risk if invalidated, then score.
- For Dependencies: likelihood = likelihood of dependency NOT being met on time.
- Only flag [act now] for score >= 16 or where mitigation lead-time is longer than time available.
- Don't invent context. If a risk needs more info to score, output [needs more info: <what>].

Paste your RAID items (one per line) + optional project context

⚡ powered by Cloudflare Workers AI · quota deducted on success

// output

// sample output

RAID — ranked by Impact × Likelihood

1. [DEPENDENCY] Compliance sign-off blocking pilot launch — score: 20/25
   Impact (1-5): 5 · Likelihood (1-5): 4
   Why: Compliance is on the critical path 4 weeks out; their review window is typically 3-4 weeks. No path to launch without it.
   Next: Owner books a 30-min review with compliance lead this week. Drafts a 2-page summary they can pre-read.
   [act now]

2. [RISK] Vendor SLA penalty if migration slips past June 30 — score: 20/25
   Impact (1-5): 5 · Likelihood (1-5): 4
   Why: Vendor penalty is contractual; 4-week pilot window is the only buffer. Mid-project slips compound.
   Next: Negotiate June 30 → July 15 ceiling NOW while you still have leverage. Backup: identify the single deferrable scope item that gives back 2 weeks.
   [act now]

3. [ISSUE] Test env has been down for 5 days — score: 16/25
   Impact (1-5): 4 · Likelihood (1-5): 4
   Why: Already happening (lik = 4). Blocks every test cycle. Pilot needs minimum 2 full test passes.
   Next: Today: escalate to platform team with a 24h SLA. If still down by Friday: book an emergency standup with platform lead.
   [act now]

4. [RISK] Two senior engineers in same parental leave window — score: 9/25
   Impact (1-5): 3 · Likelihood (1-5): 3
   Why: Known dates so likelihood is "certain" within the window. Impact bounded by what they own.
   Next: Pair-program their owned modules with a backup engineer in the 2 weeks before leave starts. Document handover.

5. [ASSUMPTION] New auth service supports our existing JWT format — score: 6/25
   Impact (1-5): 3 · Likelihood (1-5): 2
   Why: Vendor docs typically state JWT compatibility. Mitigation is a quick test.
   Next: Half-day spike — auth team writes a JWT through the new service. Output: confirmation or specific incompatibility.

// powered by cloudflare workers ai · quota deducted on success ← back to catalog