Efficiency Layer — Next Signals for Reopening Expansion

Reference documentation for Efficiency Layer — Next Signals for Reopening Expansion in Adaptive Skills.

Purpose

Explain when it is healthy to reopen the Efficiency Layer for new changes after the first trio has already been published, documented, and reinforced by review.

This guide exists to protect the repository from growing by inertia.

Current position

The repository already has:

• a roadmap
• three published efficiency skills
• a pilot kit
• a field reference
• a trio-patterns guide
• Evolution Layer evidence that says the trio is sufficient for now

That means the default posture is:

• keep using the trio
• record new evidence
• do not add a fourth skill unless a real signal appears

Good signals

Reopen the Efficiency Layer only when at least one of these happens in real work.

1. Repeated waste that the trio does not solve

Examples:

• work is still too heavy even after chunking
• resumptions still require large context reloads even with handoff notes
• teams keep continuing past healthy checkpoints because the current review pattern is too weak

2. Same pain across more than one lane or project

One isolated case is usually not enough.

A stronger signal is:

• the same operational pain appears in multiple lanes
• or the same pain appears in more than one consuming project

3. A recurring gap that does not fit existing boundaries

Examples:

• the problem is not planning, not chunking, not checkpointing, and not handoff
• the team can clearly explain why the current trio is adjacent but insufficient
• the missing move can be described without collapsing into AletheIA macro work

4. Evolution Layer evidence converges on the same conclusion

This is the strongest signal.

If multiple observations and a review all point to the same missing move, then expansion may be justified.

Weak signals

These are not good enough on their own.

• a vague feeling that the library could do more
• one difficult session
• one long task with unclear ownership
• desire to cover every productivity concern up front
• pressure to add a skill just because the trio is complete

Preferred response order

Before adding a new skill, check this order:

1. reinforced
2. no-change
3. improve a local pilot note
4. add another observation
5. add a review
6. refine an existing sidecar or example
7. only then consider a new skill

Candidate reopening areas

If expansion does become justified, the next candidates remain:

• context-compression
• prompt-framing
• model-routing

But all three should stay deferred until the signal is stronger than documentation alone.

Boundary with AletheIA

Do not reopen the Efficiency Layer for problems that are really about:

• macro framing
• escalation posture
• cross-boundary continuity policy
• gate/review ownership

Those belong to AletheIA or local operating overlays.

Recommended action when a signal appears

When a real signal appears:

1. write an observation
2. classify whether the result is reinforced, no-change, or a true gap
3. add or update a review
4. only then decide whether a proposal or new skill is justified

Healthy default

Until then, the healthy default is simple:

• keep the trio bounded
• prefer evidence over expansion
• let field use decide when the next move is real