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/zfinix/aster/agents-mdgit clone --depth 1 https://github.com/Zfinix/asterWhat 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 | $0.02676 | $0.02676 |
| Opus 5 | $0.01338 | $0.01338 |
| Sonnet 5 | $0.00535 | $0.00535 |
| Haiku 4.5 | $0.00268 | $0.00268 |
Grade A, and why
aster 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 — 238 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Aster
Rust workspace, edition 2024, stable toolchain. Crates live in crates/ and are
prefixed aster- (the exception is symbol-extractor). The crates/aster-cli
crate is the binary; everything else is a library.
Human-facing setup and PR etiquette live in CONTRIBUTING.md. This file is the
working agreement for agents editing this repo.
Global rules
- No useless comments in code.
- Match the project's existing style and patterns before writing anything new.
- No em dashes (—) in any writing.
- Never add
Co-Authored-By: Claude/Cursortrailers to commits or PRs, and strip them if found. - Use Conventional Commits for every message.
User-facing copy
Every string a user sees (TUI cells, panel cards, error boxes, status lines) is written in plain, relatable language, never harness vocabulary. "Turns", "verdict", "judge", "budget exhausted", "rate limited", "context window" are internal words; users get "tries", "check", "too many requests", "stopped after 20 tries". Errors always say what happened and what to do next in one plain sentence, with the raw provider detail demoted below. Protocol field names, config keys, env vars, and docs stay technical; only rendered copy is covered by this rule.
Rust conventions
- Read the crate you are touching and follow its existing patterns before introducing a new one.
- Inline variables into
format!braces:format!("{name}"), notformat!("{}", name). - Collapse nested
ifstatements. - Prefer method references over closures:
.map(str::trim), not.map(|s| s.trim()). - Avoid bare
booland ambiguousOptionparameters that force callers to writefoo(false)orbar(None). Prefer enums, newtypes, or named methods so the callsite reads on its own.aster_policy::ModeandPermissionModeArgare the shape to copy. - Make
matchstatements exhaustive. Avoid wildcard arms over enums we own, so adding a variant produces a compile error instead of silent fallthrough. - New traits get a doc comment explaining their role and what implementors are
expected to do.
McpInvokerincrates/aster-mcp/src/lib.rsis the example. - Prefer private modules with an explicit
pubcrate API. Keep the exported surface as small as the callers need. - Do not create helper functions that are called exactly once.
- Keep the core dependency-light. No vector DB, no external services in the review path. A new heavy dependency needs a written justification in the PR.
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 · 238 lines · 2,676 tokens per session scan A 8e8f7cdfc2a6
aster AGENTS.md is an instructions file published in the GitHub repository Zfinix/aster (40 stars, last pushed 7d ago), licensed Apache-2.0. It adds 2,676 tokens to every session, about $0.0134 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
har AGENTS.md
Instructions for os-factory/har, covering har — agent development guide, har / agent environment, harnesses in this repo, harness workflow (dogfooding) and run history.
opencode-plusplus AGENTS.md
Instructions for whut09/opencode-plusplus, covering agent guide, generated agent guide, must-read rules, default workflow and project entrypoints.
lemma-platform AGENTS.md
AGENTS.md instructions for lemma-work/lemma-platform, covering working in this repository, what is here, running anything in the backend, read these before changing behaviour and the five that get broken most.
lemma-platform CLAUDE.md
Claude Code instructions for lemma-work/lemma-platform, covering working in this repository and run python through uv.
spec-manager CLAUDE.md
Instructions for loki-ai-ch/spec-manager, covering claude.md, spec-driven development — no vibe coding, build & test commands, architecture and source layout.
spec-manager AGENTS.md
Instructions for loki-ai-ch/spec-manager, covering spec-manager workflow capsule, mandatory workflow and common commands.