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.
git clone --depth 1 https://github.com/Kulaxyz/self-learning-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/rules/kulaxyz/self-learning-skills/self-learning)<a href="https://agentmods.dev/rules/kulaxyz/self-learning-skills/self-learning"><img src="https://agentmods.dev/badge/rules/kulaxyz/self-learning-skills/self-learning/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/rules/kulaxyz/self-learning-skills/self-learning"><img src="https://agentmods.dev/badge/rules/kulaxyz/self-learning-skills/self-learning.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What 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 5.1 | $0.00637 | $0.00637 |
| Opus 5.5 | $0.00255 | $0.00255 |
| Sonnet 5.5 | $0.00127 | $0.00127 |
| Haiku 4.5 | $0.00064 | $0.00064 |
Grade A, and why
self-learning 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 1mo ago.
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 — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Self-learning
Make yourself self-improving: when you earn a reusable golden path, capture it so the next session starts knowing it instead of rediscovering it.
Recognize the moment
- A task only worked after several attempts, wrong turns, or a correction.
- You found project facts you didn't know up front: where creds/env vars live, a non-obvious command, a required sequence, a gotcha that defies the obvious.
- An operational workflow likely to recur — reach the dev/prod DB, deploy, run migrations, seed data, verify a change live.
- The user says "remember this" / "don't make me re-explain this next time".
Act on the cue immediately — don't ask permission first. Then tell the user what you saved and where.
Capture it as a rule
Write a new rule at .cursor/rules/learned/<short-name>.mdc:
---
description: <what it does AND when to use it — this is how Cursor reloads it later>
# or, if path-specific instead of description-triggered:
# globs: ["path/glob/**"]
---
# <title>
**Goal:** <the recurring task this serves>
## Steps
1. <exact command / path; keep the required order if it matters>
## Gotchas
- <non-obvious fact; where a secret lives — never the value>
## What didn't work
- <approach you ruled out, and why>
Capture the procedure (commands, paths, order, gotchas), not a one-off answer, plus the failures so next time skips the dead-ends.
Promotion rule
Only write a rule when all three hold, so guesses don't get enshrined:
- A passing check — the path was actually verified (test passed, clean exit, green build, reproduced repro). "Seemed to work" doesn't count.
- A named failure pattern — you can name the failure it avoids or diagnoses.
- At least one ruled-out dead-end — an approach you tried and eliminated.
If any is missing, leave a tentative note in project memory instead, or skip it.
Rules
- Never write secret values (tokens, passwords, connection strings, keys). Record only where they live (env var name, config file, secret manager).
- A one-line fact → put it in project notes/memory, not a whole rule.
- A genuine one-off → skip it.
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.
- 1mo ago First seen · 68 lines · 637 tokens per session scan A b344f5d4c62f
self-learning is a cursor rule published in the GitHub repository Kulaxyz/self-learning-skills (968 stars, last pushed 1mo ago), licensed MIT. It adds 637 tokens to every session, about $0.0025 per session on Opus 5.5. 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 2026-08-30.
Other cursor rules, from other repositories
ponytail
Ponytail, lazy senior dev mode. Always pick the simplest solution that works.
angular-20
This rule provides comprehensive best practices and coding standards for Angular development, focusing on modern TypeScript, standalone components, signals, and performance optimizations.
dev-standard
Apache Superset development standards and guidelines for Cursor IDE.
prefer-assertions-over-defensive-checks
Prefer assertions over defensive checks when data is guaranteed to be valid.
prefer-direct-imports-over-module-mocks
Prefer extracting a testable core over vi.mock / vi.resetModules when unit tests need to reach production logic entangled with config, env, or singletons.
config-validation-and-normalization
Config validation and normalization responsibilities (defineConfig normalizes, the orm section schema validates and resolves paths, the loader reports per-section diagnostics).