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/ashlrai/ashlrcode/claude-mdgit clone --depth 1 https://github.com/ashlrai/ashlrcodeWrote 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/ashlrai/ashlrcode/claude-md)<a href="https://agentmods.dev/instructions/ashlrai/ashlrcode/claude-md"><img src="https://agentmods.dev/badge/instructions/ashlrai/ashlrcode/claude-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 | $0.03262 | $0.03262 |
| Opus 5 | $0.01631 | $0.01631 |
| Sonnet 5 | $0.00652 | $0.00652 |
| Haiku 4.5 | $0.00326 | $0.00326 |
Grade A, and why
ashlrcode CLAUDE.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 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.
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 — 217 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AshlrCode v2.1
Multi-provider AI coding agent CLI. Open-source (MIT), npm publish-ready. 45+ tools, 18 skills, 39 built-in commands, 6 providers, 689 tests. Ink-based terminal UI with buddy system.
v2.0 Additions
- Verification Agent (
src/agent/verification.ts): Auto-validates multi-file changes via read-only sub-agent. Manual/verifycommand. Auto-suggest after 2+ file edits. - Coordinator Mode (
src/agent/coordinator.ts): Multi-agent orchestration — plans subtasks, dispatches to team members in parallel waves, auto-verifies./coordinatecommand. - LLM Dream Consolidation (
src/agent/dream.ts): Dreams are now LLM-summarized (not raw excerpts). Overlapping dreams auto-merged. Uses router on exit. - KAIROS Push Notifications (
src/agent/kairos.ts): macOS notifications when autonomous work completes or errors while user is away. - Provider-Aware Prompt Budget: System prompt assembly uses 5% of provider context limit (not static 8000). xAI gets ~100K, Ollama gets ~1.6K.
- Cost Budgeting (
src/providers/cost-tracker.ts): Budget warnings at 75%/90%/100%. Set via--max-cost. - Tool Execution Metrics (
src/agent/tool-executor.ts): Cumulative timing, success rates, avg duration per tool./statscommand. - Speculation Stats: Cache hit rate shown in turn separator (
⚡XX% cache). - 12 Model Patches: Expanded from 7 to 12 — specific patches for Llama3, CodeLlama, Mistral, DeepSeek Coder, Qwen, small models.
- Ollama Context Limits: Provider-aware (32K default) + 12 free pricing entries for common local models.
v2.1 Additions
- cmux Integration (
src/cmux/): Socket client for cmux terminal app JSON-RPC API. 9 lifecycle hooks (session start/end, agent idle, needs input, tool start/end, prompt submit, notify, error). Split spawning dispatches sub-agents to visible cmux panes. Auto-detects via$CMUX_SOCKET_PATH, all hooks are no-ops outside cmux. - Command Registry (
src/commands/): Centralized command system replacing inline switch statement. 6 command files withCommandContextinterface, auto-generated autocomplete, and category-based help grouping. - Checkpoint Workflows (
src/agent/checkpoint.ts): Coordinator can pause at checkpoint tasks. State serialized to~/.ashlrcode/checkpoints/<id>.json. Resume with/coordinate resume <id>, list with/coordinate list. Auto-cleanup after 7 days. - Skill Marketplace (
src/skills/marketplace.ts,src/skills/validator.ts): Install/update/remove skill packages from registry or URL. Git clone and tarball extraction. Skill validation (frontmatter schema, trigger conflicts, size limits). Commands:/skills install,/skills search,/skills update,/skills remove,/skills info.
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 · 217 lines · 3,262 tokens per session scan A 4c19c131ca24
ashlrcode CLAUDE.md is an instructions file published in the GitHub repository ashlrai/ashlrcode (5 stars, last pushed 1mo ago), licensed MIT. It adds 3,262 tokens to every session, about $0.0163 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.
Other instructions, from other repositories
grok-cli AGENTS.md
Instructions for superagent-ai/grok-cli, covering agents.md, cursor cloud specific instructions, quick reference, known issues and environment.
pi-xai-oauth AGENTS.md
AGENTS.md instructions for BlockedPath/pi-xai-oauth, covering agents.md — ai agent operations manual for pi-xai-oauth, project overview, key commands (exact, copy-paste ready), architecture & boundaries (must / must not) and file structure & wayfinding.
grok-subagent AGENTS.md
Instructions for Walvez/grok-subagent, covering agents.md, 项目定位, 怎么跑起来, 依赖：node.js 22+、已登录的官方 grok cli（/.grok/bin/grok 或 grok） and 安装后必须新建 codex 任务，旧会话不会热加载 skill/mcp.
grok-build-mcp AGENTS.md
AGENTS.md instructions for zai-one/grok-build-mcp, covering agents.md — grok-delegate / unofficial grok mcp, язык, typed tools vs navigator, откуда берётся корень and проект должен сам включить мост.
grok-MCP AGENTS.md
AGENTS.md instructions for nympheon/grok-MCP, covering grokmcp 维护指南, 不可破坏的运行契约, 源码职责, secret 与状态 and 登录与刷新行为.
grok-build-mcp CLAUDE.md
Claude Code instructions for zai-one/grok-build-mcp: Неофициальный MCP-мост к Grok CLI. Правила проекта живут в AGENTS.md — он импортируется ниже, отдельно открывать не нужно. Плюс skill grok-mcp как роутер по инструментам.