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/ciferateam/gitim/agents-mdgit clone --depth 1 https://github.com/CiferaTeam/GitIMWhat 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 | $0.11484 | $0.11484 |
| Opus 5 | $0.05742 | $0.05742 |
| Sonnet 5 | $0.02297 | $0.02297 |
| Haiku 4.5 | $0.01148 | $0.01148 |
Grade D, and why
GitIM AGENTS.md scanned grade D with 3 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
`$workspace/.gitim-runtime/config.json`（chmod 0600，unix 唯一权限模型）： Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
- **local → github 迁移**：需 rm -rf workspace 重建 Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- Daemon 走 `AuthData::GitHub` 分支自己推断身份（curl `/user`） How it starts
The opening of the file, as written. The whole thing — 286 lines — stays where its author put it; the contents beside it link to each section on GitHub.
GitIM
面向 Agent 团队的 AI 原生 IM 协议。纯文本文件 + Git。
架构
- 消息是
.thread文件中的行，前缀格式：[L<行号>][P<父行号>][@<handler>][<时间戳>] <正文> - 通过
P字段实现线程链 — 无需 thread_id - 续行：下一行没有
[L...]开头即为当前消息的续行 - 用户：
users/<handler>.meta.yaml，handler = GitHub handle（小写） - 技术栈：Rust daemon（核心引擎）+ Rust CLI + React WebUI
- 通信：Unix socket（默认）+ HTTP（调试模式 & WebUI）
- Git 负责持久化、同步和审计追踪
- 合规性：daemon 写入验证（主防线）+ 读取检测（第二防线）
Crate 地图
gitim-cli ──→ gitim-client ──→ [Unix socket IPC] ──→ gitim-daemon
├── gitim-core（类型 + 解析）
└── gitim-sync（git 同步）
gitim-runtime ──→ gitim-client
──→ gitim-agent-provider
├── claude / codex（原生 CLI 集成）
├── hermes / kimi（ACP 集成）
├── opencode / pi / cursor / gemini / openclaw
└── mock（测试用）
核心 crate
| Crate | 职责 | 关键模块 |
|---|---|---|
gitim-core |
数据类型、消息解析、格式化、校验 | types, parser, formatter, validator, dm, mention, link |
gitim-daemon |
主服务进程，处理所有 IM 操作 | handlers（消息/频道）, board_handlers（看板）, onboard（用户注册）, identity（身份推断）, http（SSE 推送）, state（共享状态） |
gitim-sync |
Git 同步循环、冲突解决、行号重编 | git（GitStorage 封装）, sync_loop, conflict, renumber, watcher |
gitim-client |
IPC 客户端库，封装 daemon 通信 | GitimClient（所有 API 方法）, daemon（进程管理） |
gitim-cli |
命令行工具（clap） | send, read, channels, create-channel, join-channel, status 等 |
gitim-wasm |
core/sync 纯逻辑的 wasm-bindgen 绑定，daemon-web 的协议逻辑单源 | 单文件 lib.rs；产物 pkg/ 提交进 git，改 core/sync 协议逻辑后跑 npm run build:wasm（frontend）重建并 commit |
gitim-updater |
self-update 共享核心（CLI 和 runtime 共用） | 版本解析、平台检测、fetch_latest_tag、install_update、原子替换 replace_binaries |
Agent 运行时
| Crate | 职责 | 关键模块 |
|---|---|---|
gitim-runtime |
Agent 生命周期管理、polling、HTTP API、CLI 子命令 | agent（provision）, agent_loop（消息检测 → AI 处理 → 回复）, poller, preflight, http（WebUI API）, cli（agent-facing 命令工具 → HTTP thin wrapper） |
gitim-agent-provider |
AI 提供商抽象层 | claude, codex, hermes, kimi, opencode, pi, cursor, gemini, openclaw, mock |
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.
- 2d ago First seen · 286 lines · 11,484 tokens per session scan D a556c4add46e
GitIM AGENTS.md is an instructions file published in the GitHub repository CiferaTeam/GitIM (24 stars, last pushed 4d ago), licensed Apache-2.0. It adds 11,484 tokens to every session, about $0.0574 per session on Opus 5. A static security scan graded it D with 3 findings (asks for root, recursive force delete, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
buildNext
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
spec-kit AGENTS.md
Instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
langchain AGENTS.md
Instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.