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/ekhodzitsky/kimi-wire/agents-mdgit clone --depth 1 https://github.com/ekhodzitsky/kimi-wireWrote 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/instructions/ekhodzitsky/kimi-wire/agents-md)<a href="https://agentmods.dev/instructions/ekhodzitsky/kimi-wire/agents-md"><img src="https://agentmods.dev/badge/instructions/ekhodzitsky/kimi-wire/agents-md.svg" alt="Measured on agentmods" 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 | $0.02238 | $0.02238 |
| Opus 5 | $0.01119 | $0.01119 |
| Sonnet 5 | $0.00448 | $0.00448 |
| Haiku 4.5 | $0.00224 | $0.00224 |
Grade B, and why
kimi-wire AGENTS.md scanned grade B with 1 finding 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 5d 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.
Strips warnings and disclaimersmediumAnti-refusal
Omitting safety caveats hides risk from the user and is a common jailbreak preamble.
# Documentation (must build without warnings) How it starts
The opening of the file, as written. The whole thing — 207 lines — stays where its author put it; the contents beside it link to each section on GitHub.
kimi-wire Agent Guide
This file contains agent-level conventions for the kimi-wire crate — a
standalone Rust library for the Kimi Code CLI Wire Protocol.
Contents
- Meta Principle
- Behavioral Guidelines
- Library Contract Rules
- Rust Safety Rules
- Error Handling Doctrine
- Wire Protocol Compatibility
- Build & Test
- Clippy Lint Policy
- Editing Rules
Meta Principle
Before applying any rule or refactor, ask: what problem does this solve? A newtype, a refactor, or an abstraction is justified only if it prevents a concrete bug, clarifies an invariant, or removes a footgun. If the answer is "it looks better" — revert. Decoration is not engineering.
Behavioral Guidelines
- State assumptions explicitly. If uncertain, ask before implementing.
- Minimum code. No speculative abstractions. No features beyond the request. If you write 200 lines and it could be 50, rewrite.
- Surgical changes only. Touch only what you must. Match existing style. Remove imports/variables made unused by your changes; don't delete pre-existing dead code unless asked.
- Goal-driven execution. Every task needs verifiable success criteria and a brief plan
(
Step → verify: check). - Prefer
?overunwrap/expecteven in tests where it keeps the test readable.
Library Contract Rules (Hard Constraints)
These rules protect the crate's most fragile contracts: the public API surface, serde correctness, and Wire protocol compatibility.
- Public API is opt-in. Prefer
pub(crate). Newpubitems require a concrete external caller or a paved-path rationale, plus a test. - Protocol facts must not go stale. If event names, request names, or method
parameters change, update
README.md,CHANGELOG.md, and tests in the same change set. - Serde roundtrip is a hard contract. Any change to
Event,Request,ContentPart,DisplayBlock, or method params/result types must include a serde roundtrip test. The envelope format ({"type":"...","payload":{...}}) must be preserved. - Dependencies are architecture changes. No new crate without a rationale: why std/local code is not enough, transitive impact, MSRV, license, and feature-flag consequences. Small helper crates are rejected by default.
- Refactors isolate mechanics from behavior. File moves, splits, renames, and formatting-only changes must be separate from semantic changes whenever practical.
- MSRV is 1.80. Do not use language or std features introduced after Rust 1.80.
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.
- 5d ago First seen · 207 lines · 2,238 tokens per session scan B 98d56235fcfa
kimi-wire AGENTS.md is an instructions file published in the GitHub repository ekhodzitsky/kimi-wire (2 stars, last pushed 20d ago), licensed MIT. It adds 2,238 tokens to every session, about $0.0112 per session on Opus 5. A static security scan graded it B with 1 finding (strips warnings and disclaimers). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
openpencil CLAUDE.md
Claude Code instructions for ZSeven-W/openpencil, covering claude.md, commands, architecture, code style and git commit convention.
intelligent-terminal rust-wta.instructions.md
File-scoped conventions for the WTA Rust crate.
intelligent-terminal rust.instructions.md
Concise Rust coding conventions for this repository.
anda AGENTS.md
AGENTS.md instructions for ldclabs/anda, covering agents.md, scope, repository shape, working rules and runtime boundaries.
Symbiont AGENTS.md
AGENTS.md instructions for ThirdKeyAI/Symbiont, covering symbiont — agent instructions, project structure, build and test, code style and commit guidelines.
openinterpreter AGENTS.md
Instructions for openinterpreter/openinterpreter, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.