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 instructions/voleeo/voleeo-api/agents-mdgit clone --depth 1 https://github.com/voleeo/voleeo-apiWhat 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 | $0.06134 | $0.06134 |
| Opus 5 | $0.03067 | $0.03067 |
| Sonnet 5 | $0.01227 | $0.01227 |
| Haiku 4.5 | $0.00613 | $0.00613 |
Grade A, and why
voleeo-api AGENTS.md 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 3d 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 — 253 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Guidance for AI coding agents working in this repository.
File-size rule: this document must stay under 300 lines. When you add a new rule or section, tighten or remove existing content first — never just append. A bloated guide is read less, not more. Keep prose terse, examples one-line, and prefer references to inline duplication.
Product vision
Voleeo is an AI-native local toolkit with MCP bridge. Every feature serves one goal: an AI agent (via MCP) and a developer share the same view of the system. UI, data models, and IPC are designed for both human and machine consumers.
Model routing (delegated sub-work only)
Applies to subagents/workflows spawned via the Agent/Workflow tools — pick model per task. The main session model is user-controlled (/model) and never auto-switches; this table only routes delegated work. When unsure, step up a tier.
- Haiku — mechanical, fully-specified: keyword/file searches, fan-out reads (
Explore), renames, formatting, log/grep scans, applying a known edit across files. Also commit messages and PR descriptions (use the cheapest tier). - Sonnet — default for real work: feature implementation, code review, writing tests, focused single-file/single-bug debugging, doc edits.
- Opus — hard reasoning: architecture/design, cross-crate refactors, render-loop/concurrency/IPC-shape bugs (see numbered rules), security review, ambiguous multi-step tasks, adversarial verification of findings.
Tool discipline
- Locate before reading. Use grep/find to confirm a file holds what you need before
Read; never read whole directories. - Budget the work. Aim to finish a task in under 20 tool calls. After ~10 calls with no visible progress, stop and explain the blocker rather than pushing on.
Code quality
KISS · DRY · YAGNI. When planning and writing code, default to the simplest thing that works: stdlib or an already-installed dep before new code, one line before fifty, a native/CSS/DB feature before app code. Build nothing for a need that isn't here yet. Factor real duplication — never abstract speculatively (no interface with one impl, no config for a constant). Shortest working diff wins; if you can't name why a piece of code must exist, cut 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.
- 3d ago First seen · 253 lines · 6,134 tokens per session scan A 1c4345c4472b
voleeo-api AGENTS.md is an instructions file published in the GitHub repository voleeo/voleeo-api (117 stars, last pushed 11d ago), licensed MIT. It adds 6,134 tokens to every session, about $0.0307 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 2026-08-30.
Other instructions, from other repositories
KtorMonitor AGENTS.md
Instructions for CosminMihuMDC/KtorMonitor, covering agents.md — ktormonitor, 1. project overview, 2. module structure, source-set layout (core) and 3. build, test, run.
monkey AGENTS.md
Instructions for monkey/monkey, covering agents, repository map, main runtime flow, build and verification and commit style used in this repository.
casbin-gateway CLAUDE.md
Claude Code instructions for apache/casbin-gateway, covering claude.md and code style.
httpstat AGENTS.md
Instructions for reorx/httpstat, covering httpstat 项目规划, 1) 项目现状, 已完成的现代化工作, 2) 剩余技术债 and a. 测试可靠性不足（高优先级）.
igniter-js GEMINI.md
Gemini CLI instructions for felipebarcelospro/igniter-js, covering 1. identity and profile, 2. about the igniter.js monorepo, 3. personality and communication, 4. lia's core responsibilities (the 4 pillars) and 5. technical guidelines and methodology.
igniter-js AGENTS.md
AGENTS.md instructions for felipebarcelospro/igniter-js, covering lia - ai agent for igniter.js, 1. identity & mission, core mission, key responsibilities and 2. project overview.