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/totocaster/arrowhead/agents-mdgit clone --depth 1 https://github.com/totocaster/arrowheadWhat 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 | $0.00755 | $0.00755 |
| Opus 5 | $0.00378 | $0.00378 |
| Sonnet 5 | $0.00151 | $0.00151 |
| Haiku 4.5 | $0.00076 | $0.00076 |
Grade A, and why
arrowhead 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 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.
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 — 47 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Arrowhead Coding Agent Playbook
Mission
- Deliver the Arrowhead Rust rewrite according to the consolidated spec in
README.md. - Preserve repository cleanliness and avoid ambiguous states; never leave
todo!()panics in committed code. - Prioritise maintainability, clear error handling, and strong test coverage using the provided fixture vault.
Ground Rules
- Toolchain: Build against Rust 1.86 (2024 edition). Update
rust-toolchain.tomlif a later MSRV is required; call it out in reviews. - Features: Semantic search ships by default using sqlite-vec; keep embeddings healthy and avoid regressing the always-on pipeline.
- Style: Follow idiomatic Rust (clippy clean,
cargo fmt). Prefer explicit structs/enums over loose maps; document non-obvious flows with concise comments. - Error Handling: Use
anyhow/thiserroras scoped in the spec. Return actionable errors instead of panicking. - Testing: Add unit tests alongside code (
#[cfg(test)]). Integration tests undertests/integration/will land later; plan your APIs so those paths stay testable. Usetests/fixtures/test-vaultas read-only input; write indexes to temp dirs. - Docs: Update specs/design docs when behaviour or architecture shifts so the expectations stay current.
Workflow Expectations
- Start complex tasks with a brief plan (2–5 steps max) and keep it up to date.
- For MCP tooling, call
mcp.discovery.get_vault_conventionsright after initialize and before any note creation, update, or deletion so agents respect vault naming rules. - Stage work incrementally; validate with
cargo fmt,cargo check,cargo clippy --all-targets -- -D warnings, and relevant tests before surfacing results or pushing. - Treat those commands as the “pre-flight” checklist before every push/PR. Run
cargo fmt --all(not just--check), rebuild, and rerun tests after the final file edit so CI never redlines on formatting or missing coverage. - Summaries must highlight behavioural changes, tests executed, and next steps. Flag known gaps or follow-up items.
- If external dependencies require network or MSRV bumps, pause and confirm with the designer before proceeding.
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 · 47 lines · 755 tokens per session scan A d7e5f27ba769
arrowhead AGENTS.md is an instructions file published in the GitHub repository totocaster/arrowhead (43 stars, last pushed 21d ago), licensed MIT. It adds 755 tokens to every session, about $0.0038 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
mcp-server-excel coverage-prevention-strategy.instructions.md
Instructions for sbroenne/mcp-server-excel, covering generated surface coverage, contract change workflow, required checks and common incomplete changes.
my-claude-code-setup GEMINI.md
Instructions for centminmod/my-claude-code-setup, covering gemini.md, ai guidance, memory bank system, core context files and project overview.
mcp-server-excel mcp-llm-guidance.instructions.md
Instructions for sbroenne/mcp-server-excel, a project described as: Automate real Microsoft Excel with AI via MCP Server or CLI — Power Query, DAX, VBA, PivotTables, charts, and 326 operations.
cship CLAUDE.md
Instructions for stephenleo/cship, covering cship — ai agent instructions, environment quirks, non-negotiable code patterns, before submitting a pr and project structure.
agentic-workflow CLAUDE.md
Instructions for gtrabanco/agentic-workflow, covering claude.md, repository layout, working rules, authoring a skill and hand off, don't compose across a model/effort boundary.
Claude-Code-Promts-Skills CLAUDE.md
Instructions for Rtur2003/Claude-Code-Promts-Skills, covering claude.md, project overview, repository structure, file naming conventions and prompt structure standards.