Efficiency Layer First Pilot Guide

Reference documentation for Efficiency Layer First Pilot Guide in Adaptive Skills.

Goal

Run the first real pilot of the Efficiency Layer without turning it into a broad productivity program.

The purpose is to answer:

• do the first three efficiency skills reduce avoidable operational waste?
• do they make bounded work easier to continue, pause, or stop?
• do their boundaries stay clear relative to AletheIA and the other skills?

Recommended pilot shape

Keep the pilot intentionally small:

• 1 project or lane
• 1 to 3 recurring task shapes
• the initial efficiency trio only:
  • task-chunking
  • handoff-summary
  • checkpoint-review
• 1 to 2 weeks of real use

Do not start with a broad process rollout.

Good pilot lanes

Choose a lane where:

• work tends to arrive oversized
• context grows noticeably across rounds
• handoffs or restarts are common
• the team can compare before and after with real work

Good examples:

• a feature lane with repeated slice creep
• a workstream with many pauses and resumptions
• a debugging or delivery loop that keeps expanding past one clean round

What to keep local

The pilot should not pull in:

• local release gates
• local review ownership maps
• org-specific scheduling or staffing rules
• company-specific model policy

Those remain local overlays or AletheIA-level macro posture.

Reference case

A first-validation reference case (Crisis Monitor) is available in docs/efficiency-layer-crisis-monitor-reference.md. Use it when you want one concrete field example of the trio mapped onto a real product lane without treating that product repo as the implementation target. Per ADR-002, this is the first concrete instance, not the canonical pattern.

Suggested pilot loop

1. Pick one recurring lane

Write down:

• project
• lane
• the recurring pain pattern
• which of the three efficiency skills are in scope

2. Use only the trio

Prefer the smallest fit:

• task-chunking when the task arrives too large
• checkpoint-review when a round needs a stop-or-continue decision
• handoff-summary when the round should close and resume later

3. Capture one short note per meaningful use

Track:

• which skill was used
• why it was triggered
• whether it made the next step clearer
• whether it felt lighter or heavier than the previous way of working

4. Review after a small number of rounds

At the end of the pilot, decide:

• which skill was genuinely useful
• which boundary was still confusing
• whether another efficiency skill is justified
• whether the trio is already enough for the lane

Success criteria

The first Efficiency Layer pilot is successful if:

• at least one recurring lane feels easier to keep bounded
• stop-or-continue decisions become more explicit
• handoffs become smaller and more resumable
• the trio stays clearly separate from AletheIA macro work

Failure signals

• the pilot becomes a generic productivity initiative
• the trio is used everywhere instead of where pain is real
• the team cannot tell when a problem belongs to feature planning versus chunking
• checkpoint or handoff notes become long ceremony instead of useful compression

Output artifact

Use templates/pilot/efficiency-pilot-report.md to capture the result.