agentleFS
Sign inStart free
Fadi88Fadi88/one-man-startup/.cursor/rules/skill-context-budget.mdc

one-man-startup / rules

Cursor rule16 starsChanged 2 days agoMITView on GitHub

How to spend context - open what the task names, grep before reading, read ranges, prefer a checker's output, say what was opened. Use on every task; it is what keeps a long change from losing its own thread halfway.

We split this skill-context-budget.mdc into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Every file read costs tokens, money and, worst, the thread of the work when the context fills and gets compacted halfway through a change.

    line 10
  2. Distinctive:

    Open the file the task names,: not its neighbours.

    line 13

≈ “Common” is an estimate: at least 1 in 5 of the 6,350 Cursor rule files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
skill-context-budget.mdc
---
description: "How to spend context - open what the task names, grep before reading, read ranges, prefer a checker's output, say what was opened. Use on every task; it is what keeps a long change from losing its own thread halfway."
alwaysApply: false
---

<!-- Generated by tools/sync_adapters.py from skills/context-budget/SKILL.md. Edit the source, not this file. -->

# Context is a budget

Every file read costs tokens, money and, worst, the thread of the work when the context fills
and gets compacted halfway through a change. Spend it on purpose.

- **Open the file the task names,** not its neighbours.
- **Grep before reading.** An id, a symbol or a function name lands on the line, and the line is
  often the whole answer.
- **Read a range** of a long file, not the whole file. `docs/HISTORY.md` and the requirement set
  are never read whole.
- **Prefer a checker's output** to the source it checks.
- **Pass artifacts by path.** A hand-off names the file; it does not paste it.
- **Say what you opened** in your hand-back, so the next agent does not open it again.

This never buys tokens with accuracy. If the answer needs the file, read the file. A wrong
answer costs more than any read.

## Keep the shared files small

`AGENTS.md`, each agent file and each `SKILL.md` stay short enough to be loaded every time
without thought. Detail goes in a sibling file the skill points to. The sweep warns when one
grows past its budget.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free