agentic-loop

A repeatable process for making code changes: understand the task, find the relevant code, plan a small change, implement it, test it, and review the result. It is tailored to a cross-platform file manager with a Go backend and a React user interface.

In plain words
What is it for?
Use it for every development task in this repository, including new features, refactoring, and bug fixes. It guides code location, test placement, validation, and iteration until the task is complete.
Why use it?
It gives coding tasks a consistent path from request to verified result. It also helps keep shared file logic, the user interface, and operating-system-specific code in the right boundaries.

Cursor rule for Cursor

Install

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.

agentmods
npx agentmods add rules/arxdsilva/vault/agentic-loop
Clone the repo
git clone --depth 1 https://github.com/arxdsilva/vault

Made for: Cursor.

Per session 919 This file is loaded in full into every session.
When invoked 919 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00919 $0.00919
Opus 5 $0.00460 $0.00460
Sonnet 5 $0.00184 $0.00184
Haiku 4.5 $0.00092 $0.00092

Measured yesterday against content hash 2771cb1dcb0b, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agentic-loop 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.

.cursor/rules/agentic-loop.mdc · 99 lines

How it starts

The opening of the file, as written. The whole thing — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Vault Agentic Development Loop

Follow this loop on every development task in this repo.

Vault is a minimal, cross-platform file manager (Wails v3 Go backend + React/TypeScript/ Tailwind frontend) aiming to replace Finder (macOS) and Windows File Explorer: search-first navigation, zero chrome, keyboard-first, instant everything. Design spec: docs/SPEC.md; build progress: docs/STATUS.md.

The Loop

Repeat until the definition-of-done checklist is green:

Understand → Locate → Plan → Implement → Validate → Self-review → (iterate)

1. Understand

  • Restate the task goal and constraints (new capability vs. refactor vs. bug fix).
  • Identify whether the change touches core file-system logic, the UI/view layer, or a platform-specific adapter (see architecture.mdc).

2. Locate

  • Find existing code before writing new code.
  • For tests: co-located with the source, following whatever convention the chosen stack uses (see unit-tests-tdd.mdc).

3. Plan (minimal change)

  • Propose the smallest diff that solves the task.
  • Flag if the change crosses the core/UI boundary or touches a platform-specific adapter (macOS vs. Windows vs. Linux) — keep platform code isolated (architecture.mdc).
  • Flag if the change affects destructive file operations (delete, move, overwrite) — these need extra care and, ideally, a confirmation path and/or undo.

4. Implement

TDD first for behavior changes (see unit-tests-tdd.mdc): failing test → implement → refactor.

Area Rule
Scope Small, focused diffs; preserve existing public API/CLI/UI behavior unless asked
Architecture Core file-system logic stays UI-agnostic; platform quirks stay behind an adapter (architecture.mdc)
Safety Never perform destructive file operations without a clear, tested path; prefer trash/recycle-bin semantics over permanent delete where the OS supports it
Errors Surface file-system errors (permissions, missing paths, in-use files) to the user; don't swallow or panic
Formatting Use the formatter/linter standard for the chosen stack

Read the full file on GitHub · 99 lines

Changes

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.

  1. yesterday First seen · 99 lines · 919 tokens per session scan A 2771cb1dcb0b

Subscribe to this mod's changes

agentic-loop is a cursor rule published in the GitHub repository arxdsilva/vault (3 stars, last pushed 1mo ago), licensed MIT. It adds 919 tokens to every session, about $0.0046 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.