Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add rules/afaraha8403/balakit/commentsgit clone --depth 1 https://github.com/afaraha8403/balakitWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00564 | $0.00564 |
| Opus 5 | $0.00282 | $0.00282 |
| Sonnet 5 | $0.00113 | $0.00113 |
| Haiku 4.5 | $0.00056 | $0.00056 |
Grade A, and why
comments scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Comments & Documentation
Default to zero comments
Names document code. Add an inline comment only when the WHY is non-obvious:
- A hidden constraint not visible in the code.
- A subtle invariant a reader might violate.
- A workaround for a specific bug (link the upstream issue).
- Behavior a competent reader would find surprising.
If removing the comment wouldn't confuse a future reader, delete it.
Never write
- Restatements of the code —
increment counterabovecounter++. - Task / ticket references —
added for CAN-123,fixes billing bug. - Caller references —
used by the contacts module. - Change history —
was async before,removed old logic. - Commented-out code.
- Banner comments —
=== SECTION ===.
Authorized Tracking Tags (Exception)
You MUST use the following authorized tags to track technical debt and risks (as required by the Global Rules), using your language's comment syntax. These should cleanly describe the issue for IDE tracking:
BUG:— For critical, immediate issues.FIXME:— For broken code or technical debt needing refactor.TODO:— For general planned work.CONCERN:— For edge cases or architectural risks.OPTIMIZE:— For performance improvements or clean-up tasks.
Document every exported symbol
Every part of a module's public surface — exported functions, types, classes, constants — gets a documentation comment in your language's convention (doc comment, docstring, etc.). Internal/private helpers only get one when the WHY is non-obvious.
- The first line is a one-sentence summary ending with a period.
- Document a parameter only when its name isn't self-explanatory. Never restate information the signature already carries (e.g. the type, in a typed language).
- Document the return only when it isn't obvious from the function name.
- Document every error/exception the function can raise as part of its contract.
- Add an example for non-trivial public APIs and shared utilities.
- Don't duplicate what the language already expresses — let the type system, signatures, and tooling do their job.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 55 lines · 564 tokens per session scan A b0e7a330a110
comments is a cursor rule published in the GitHub repository afaraha8403/balakit (1 stars, last pushed 4d ago), licensed MIT. It adds 564 tokens to every session, about $0.0028 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other cursor rules, from other repositories
ui
React components must be functional and prop-typed.
skill-creator
Create, edit, evaluate, and debug the skills in this repo, including running their evals and fixing a description that fails to trigger. Use when user says 'help me build a new skill', 'add a skill for X', 'run the evals for the tf skill', 'run the behavioral evals', 'my skill is not triggering', 'fix this skill's…
java-springboot-jpa-cursorrules-prompt-file
description: "Cursor rules for Java development with Springboot and JPA integration." globs: / alwaysApply: false.
tf-plan
Review a Terraform plan before applying it: destroys and replacements of data-bearing resources, secrets readable in plan output, out-of-band drift, blast radius, and whether the apply is bound to the plan you actually reviewed. Use when user says 'review my plan', 'is this plan safe to apply', 'check tfplan', 'what…
clarify-first
Infer-and-act by default — ask only on high confusion weight, after inspecting, with a decision-linked question.
102-python-patterns
Python coding patterns and conventions.