agentsclimarketplace

Adaptive token efficiency

Skill gunwoo55/adaptive-token-efficiency

Lightweight fast-path instruction package for AI coding agents: reduces unnecessary context loading and adds state-aware validation.

Install
npx -y skills add gunwoo55/adaptive-token-efficiency

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 26 days oldThe repository was created 26 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Escalation-only workflow for materially ambiguous, research-heavy, high-impact, cross-system, or failed-validation tasks. Do not load for clear, bounded, low-risk work; use the AGENTS fast path.

SKILL.md

1.7 KB, 298 tokens by cl100k_base, as published. Nobody here has run it

Adaptive Token Efficiency v2.1

Entry

Do not load this file for a clear, bounded, low-risk task with known acceptance criteria. If loaded without a trigger, return to the fast path immediately. Never read EVALS.md or BENCHMARKS.md during normal work.

Route

Open exactly one matching module:

  • uncertain cause/requirements or failed focused validation → references/investigation.md
  • current or external evidence → references/research.md
  • migration, broad artifact, or cross-system/repository work → references/large-change.md
  • external-state or hard-to-reverse action → references/high-impact-action.md

Open a second only when two material risks genuinely interact.

Contract

Keep TARGET | ACCEPTANCE | MATERIAL RISK | NEXT CHECK.

  • Reuse available context; read only what can change the decision.
  • Make the smallest coherent action; batch related reads, writes, and checks.
  • Expand only after failed validation, material uncertainty/conflict, a new system boundary, high impact, or an explicit depth request.
  • Spend saved effort on one integrated acceptance pass, not extra exploration. For interactive changes, synchronize values, controls, status/help text, persistence, and accessibility.
  • Change a failed method once; never repeat unchanged calls or checks.
  • Return the deliverable first, with compact validation and material limits; no routine narration or private chain-of-thought.

Stop when acceptance passes.

What ships with it: 12 files

12.2 KB alongside SKILL.md, 3 of them executable

references/

tests/

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.