process-optimization

Process Optimization

Analyze existing processes and recommend improvements.

Lark-native execution (depth core: LARK-PATTERNS, LARK-RECIPES, LARK-FUSION). Map the current state from real Lark signals, not guesses: pull the existing SOP from Wiki/Docs (lark_doc_search/lark_doc_fetch, jq-projected, P3), measure approval-step waiting from Lark Approval, and quantify handoffs from Lark Task cycle times. Surface the before/after + recommendations as an interactive card (P4), spin improvement items into Lark Tasks (P2 dry_run), and update the SOP in Wiki (P8).

Analysis Framework

1. Map Current State

• Document every step, decision point, and handoff
• Identify who does what and how long each step takes
• Note manual steps, approvals, and waiting times

2. Identify Waste

• Waiting: Time spent in queues or waiting for approvals
• Rework: Steps that fail and need to be redone
• Handoffs: Each handoff is a potential point of failure or delay
• Over-processing: Steps that add no value
• Manual work: Tasks that could be automated

3. Design Future State

• Eliminate unnecessary steps
• Automate where possible
• Reduce handoffs
• Parallelize independent steps
• Add checkpoints (not gates)

4. Measure Impact

• Time saved per cycle
• Error rate reduction
• Cost savings
• Employee satisfaction improvement

Lark-native execution

Map current state from real data:

• Existing SOP: lark_doc_search (jq .data.results[].title) → lark_doc_fetch (jq .data.content[:4000]) to recover the documented steps (P3).
• Waiting/approval steps: if the process runs on a Lark Approval flow, delegate to approval-flow-sla to find the bottleneck node and per-approver dwell time — that IS the "waiting waste" measurement.
• Handoffs/rework: read the process tracker Base with lark_base_search (REQUIRES search_fields, no jq — narrow with select_fields; discover field names via lark_api GET /open-apis/bitable/v1/apps/{base}/tables/{table}/fields if unknown) or Lark Task cycle times to quantify how long items sit between handoffs.

Quantify and recommend (card, P4):

• Post the before/after comparison + top improvements as an interactive card (lark_im_card_send, print_json: true first): status pills per step (🟢 keep / 🟡 streamline / 🔴 cut), actions footer ("Apply changes" / "Open SOP").

Execute the improvement plan:

• For each improvement, resolve the owner (lark_contact_search, P1) → lark_task_create(dry_run) → confirm → commit.
• For automatable steps in a Base-driven process, delegate to base-automation to build the Base workflow/bot reminder rather than describing automation abstractly.

Update the SOP (Wiki, P8):

• Write the future-state SOP back to Wiki (lark_wiki_node_create / lark-doc skill) so the documented process matches the optimized one.

Output

Produce a before/after process comparison with specific improvement recommendations, estimated impact, and an implementation plan — measured from real Lark signals, surfaced as a card, executed as Lark Tasks, and reflected back into the Wiki SOP.