Ops Process Audit

Search vault for process context — Find references and issues:

Use the `search_memory` MCP tool with query: "<process-name> bottleneck issue problem delay"

Review meeting notes, team docs, project retrospectives mentioning this process

Search daily log for pain points — Identify recurring complaints:

grep -r "<process-name>\|bottleneck\|delay\|waiting\|stuck\|manual\|repetitive" ~/.second-brain/vault/daily/ | tail -50

Extract: when complaints happen, who's affected, impact severity

Search Gmail for complaints/escalations — Find process issues reported:

Use the `search_emails` MCP tool with gmail_query: "subject:(urgent|escalation|issue|problem|delay) <process-area>"

Extract: complaint details, who complained, frequency, resolution

Analyze current process — Document:

• Current State — Steps as currently performed (may differ from documented SOP)
• Actors — Who is involved at each step
• Bottlenecks — Where delays occur:
  • Manual handoffs between teams
  • Waiting for approvals or decisions
  • Rework or quality issues requiring iteration
  • Dependency on unavailable person/system
• Cycle Time — How long does full process take? Theoretical vs. actual?
• Cost — How much does this process cost to run? (labor, tools, opportunity cost)
• Pain Points — From email/daily log analysis, what goes wrong most often?

Calculate efficiency metrics:

• Value-Add Time = time spent on actual productive work
• Non-Value-Add Time = waiting, rework, handoffs, approvals
• Efficiency Ratio = (Value-Add / Total Time) × 100%
• Bottleneck Impact = % of total cycle time spent at biggest bottleneck
• Error Rate = % of processes requiring rework/iteration (if available from data)

Identify improvement opportunities:

• Can steps be combined or parallelized?
• Can automation replace manual work? (RPA, workflow tool, etc.)
• Are approvals/reviews necessary or just habit?
• Can bottleneck be eliminated (different owner, better tooling, etc.)?
• What would cut cycle time in half?

Generate audit report with sections:

• Executive Summary — Current state, main bottleneck, recommended fix, estimated impact
• Process Overview — Current documented process (link to SOP if exists)
• Current Performance Metrics:
  • Cycle time (baseline to target)
  • Efficiency ratio (value-add vs. non-value-add)
  • Error/rework rate
  • Cost to run process
• Bottlenecks Identified — Ranked by impact:
  • Bottleneck name, location in process, time lost, root cause
  • How often does this create problems?
• Process Issues from History:
  • Complaints/escalations from emails and daily log
  • Recurring problems with pattern of occurrence
• Improvement Recommendations — 3-5 prioritized recommendations:
  • What to change and why
  • Expected impact (time saved, cost, quality improvement)
  • Implementation effort and timeline
  • Owner/responsible team
  • Related tools/resources needed
• Implementation Plan — Detailed steps for top 1-2 recommendations
• Success Metrics — How to measure improvement once implemented

Save audit report to:

~/.second-brain/vault/ops/audits/YYYY-MM-DD-<process-slug>-audit.md

Create Asana tasks for improvement actions:

for each recommendation:
  Use the `list_tasks` MCP tool (task creation via Asana API — log the action item with `log_note` instead)
    --project "Operations" \
    --name "[Recommendation]: [brief description]" \
    --assignee "[owner]" \
    --due-date "[implementation-date]" \
    --description "[details and expected impact]"

Optionally create meeting event for process improvement kickoff:

Log the calendar item with `log_note` (calendar creation not yet in MCP)
  --title "Process Audit Review: [process-name]" \
  --date "+3 days" \
  --attendees "[stakeholders]"

Log to daily log:

- **HH:MM** — [OPS] Process audit: <process-name> — Main bottleneck: <bottleneck> — Top recommendation: <rec> (est. <impact>)