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/avala-ai/agent-code/agents-mdgit clone --depth 1 https://github.com/avala-ai/agent-codeWhat 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 | $0.02846 | $0.02846 |
| Opus 5 | $0.01423 | $0.01423 |
| Sonnet 5 | $0.00569 | $0.00569 |
| Haiku 4.5 | $0.00285 | $0.00285 |
Grade C, and why
agent-code AGENTS.md scanned grade C 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 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.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
The bash tool warns on patterns like `rm -rf`, `git reset --hard`, `DROP TABLE`, etc. and blocks writes to system paths (`/etc`, `/usr`, `/bin`, `/sbin`, `/boot`, `/sys`, `/proc`). Do not narrow these checks. If you add How it starts
The opening of the file, as written. The whole thing — 222 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Instructions for AI coding agents working in this repository. Read this file in full before making changes. It captures the non-obvious rules — conventions you cannot derive from just grepping the code.
This repo is agent-code (Avala AI): a Rust terminal coding agent, shipped as both a CLI and an embeddable library. Because this project is itself an agent, changes here affect the safety posture of every user who runs it. Move carefully.
Throughput (parallel agents / multi-crate work): read SHIP.md — blast radius, non-overlapping lanes, one git worktree per concurrent writer, atomic commits. Opt-in for multi-agent, multi-crate, or bloating PRs — not tiny single-crate edits. Does not lower the security or CI bar.
1. Repo layout
crates/
lib/ # agent-code-lib — LLM providers, tools, query loop, memory,
# permissions, MCP client, skills, compaction. The engine.
cli/ # agent-code — binary: REPL, TUI, slash commands, streaming output.
eval/ # agent-code-eval — behavioral evaluation harness (not published).
client/ # Flutter desktop/web client that talks to `agent --serve`.
packages/ # TypeScript client library (`agent_code_client`).
npm/ # npm installer wrapper.
docs/ # Mintlify + mdBook docs (architecture, guides, reference).
evals/ # Eval fixtures and scenarios consumed by `crates/eval`.
scripts/ # E2E shell harnesses.
.github/workflows/ # CI — `ci.yml` is the canonical gate.
Supporting docs you should read before a non-trivial change:
README.md, ARCHITECTURE.md, SECURITY.md, CONTRIBUTING.md, ROADMAP.md, RELEASING.md, CHANGELOG.md, SHIP.md (when multi-agent or multi-crate).
2. Build, test, lint — the CI gate
Every PR must pass ci.yml. Run these locally before pushing; they are the exact commands CI runs.
cargo check --all-targets # compile
cargo test --all-targets # unit + integration + doc tests
cargo clippy --all-targets -- -D warnings # lint — warnings are errors
cargo fmt --all -- --check # format gate
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 · 222 lines · 2,846 tokens per session scan C edd3f4b40c62
agent-code AGENTS.md is an instructions file published in the GitHub repository avala-ai/agent-code (23 stars, last pushed 7d ago), licensed MIT. It adds 2,846 tokens to every session, about $0.0142 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
openevolve CLAUDE.md
Instructions for algorithmicsuperintelligence/openevolve, covering claude.md, essential commands, development setup, install in development mode with all dependencies and or use makefile.
seektty AGENTS.md
Instructions for Hilbert-beinghappy/seektty: This repository ships one out-of-tree DeepSeek Harness Bundle. Harness remains the only owner of Agent, Session, model, settings, permissions, Profile, plugin, and persistence state.
amp-acp AGENTS.md
Instructions for tao12345666333/amp-acp, covering agents.md, commands, architecture and code style.
roamcode AGENTS.md
Instructions for burakgon/roamcode, covering repository instructions for coding agents, public-repository safety, stable release and ota contract, releasing a stable version and ota changes.
dev-workspace AGENTS.md
Instructions for iwe-org/dev-workspace, covering agent operating manual, start of every session, the operating loop, conventions and iwe basics.
ob-1 AGENTS.md
Instructions for Overbrilliant/ob-1, covering agents.md, project index, stack, commands and key files.