Run any Skill in Manus with one click

context-budget

Stars2

Forks0

UpdatedJuly 12, 2026 at 12:09

Keep the agent's context lean so accuracy doesn't collapse. Use on long sessions, big files, or when the agent starts hallucinating.

Installation

Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.

Run Skill in Manus

Source

YaoYinYing

YaoYinYing/REvoDesign

View GitHub Repository View Creator Repositories

Download

Run Skill in Manus

Related occupationsSOC

Based on SOC occupation classification

Computer Occupations, All OtherComputer and Mathematical Occupations·SOC 15-1299

SKILL.md

readonly

name	context-budget
description	Keep the agent's context lean so accuracy doesn't collapse. Use on long sessions, big files, or when the agent starts hallucinating.
when_to_use	long session, huge file dump, agent forgetting earlier steps, accuracy dropping

Context Budget

More context is not better. Past a threshold, accuracy falls off a cliff (context rot).

Don't dump whole files — read the relevant function/section, not the 2000-line module.
Summarize, don't accumulate — replace a 200K-token transcript with a 4K recap of the load-bearing facts.
Drop dead context — once a sub-task is done, its detail leaves the window. Keep the decision, drop the trace.
State on disk, not in context — progress goes to a file the next turn re-reads, not into an ever-growing prompt. A bloated standing context (CLAUDE.md) taxes every single turn. Trim weekly. If the agent is confidently wrong, suspect the context before the model.

More from this repository

same repository

adversarial-verify

YaoYinYing/REvoDesign

Review a diff against the goal spec assuming the code is BROKEN. The reviewer that lives in the maker's head always agrees with itself — this pulls review into a hostile, separate pass. Invoke after every code change before marking work done.

2026-07-122

authz-check

YaoYinYing/REvoDesign

Verify that an endpoint checks ownership, not just authentication. Use on any handler that reads or mutates user data.

2026-07-122

bisect-regression

YaoYinYing/REvoDesign

Find the exact commit that introduced a bug. Use when something worked before and broke, and you don't know which change did it.

2026-07-122

changelog-from-diff

YaoYinYing/REvoDesign

Turn a set of commits or a diff into a clean, user-facing changelog entry. Use before a release or PR description.

2026-07-122

clean-commits

YaoYinYing/REvoDesign

Turn messy WIP into clean, atomic commits with messages that explain why. Use before opening a PR.

2026-07-122

design-system

YaoYinYing/REvoDesign

Make frontend output look intentional, not AI-generated. Use for any UI work — components, pages, layouts.

2026-07-122

name	context-budget
description	Keep the agent's context lean so accuracy doesn't collapse. Use on long sessions, big files, or when the agent starts hallucinating.
when_to_use	long session, huge file dump, agent forgetting earlier steps, accuracy dropping

Context Budget

More context is not better. Past a threshold, accuracy falls off a cliff (context rot).

Don't dump whole files — read the relevant function/section, not the 2000-line module.
Summarize, don't accumulate — replace a 200K-token transcript with a 4K recap of the load-bearing facts.
Drop dead context — once a sub-task is done, its detail leaves the window. Keep the decision, drop the trace.
State on disk, not in context — progress goes to a file the next turn re-reads, not into an ever-growing prompt. A bloated standing context (CLAUDE.md) taxes every single turn. Trim weekly. If the agent is confidently wrong, suspect the context before the model.