agent_00 AGENTS.md

Repository guidance for agent_00, a Go command-line tool that runs coding agents against GitHub repositories and creates pull requests. It records the project’s scope, commands, and code organization.

In plain words
What is it for?
Use it when changing agent_00: check the project structure, follow its repository conventions, and run the documented build and test commands.
Why use it?
It gives coding agents evidence-based rules for working in the repository, reducing guesses about supported commands, required Go versions, and where code belongs.

Instructions file for CodexOpenCode

Install

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.

agentmods
npx agentmods add instructions/molten-bot/agent_00/agents-md
Clone the repo
git clone --depth 1 https://github.com/Molten-Bot/agent_00

Made for: Codex, OpenCode.

Per session 1,344 This file is loaded in full into every session.
When invoked 1,344 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.01344 $0.01344
Opus 5 $0.00672 $0.00672
Sonnet 5 $0.00269 $0.00269
Haiku 4.5 $0.00134 $0.00134

Measured 2d ago against content hash 96d7d165dd42, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agent_00 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.

AGENTS.md · 68 lines

How it starts

The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Scope

  • These instructions apply to the whole repository. A nested AGENTS.md closer to the edited file takes precedence for that subtree.
  • This file is repository-local guidance for agent_00. The separate library/AGENTS.md file is the runtime seed copied into task workspaces for agents working on customer repositories.
  • Keep agent instructions evidence-backed. Add commands, paths, and conventions only when they are present in source, tests, docs, Dockerfiles, or CI.

Project Overview

  • agent_00 is a Go module and CLI for running AI coding agents against GitHub repositories and turning prompts into pull requests.
  • The executable lives in cmd/harness. Shared behavior belongs in focused internal/* packages: app, agentruntime, config, execx, hub, library, multiplex, slug, web, and workspace.
  • The local web UI is served from embedded files under internal/web/static; there is no separate npm frontend build pipeline in this repo.

Commands

  • Use Go 1.26.5 or newer, as declared in go.mod.
  • Match CI before finishing broad changes:
    • go mod download
    • go build ./...
    • go build -o bin/harness ./cmd/harness
    • go test ./...
  • For focused validation, prefer the smallest relevant package or test first, for example go test ./internal/config or go test ./internal/app -run TestName, then broaden when the change has wider impact.
  • Build and run the local harness with go build -o bin/harness ./cmd/harness and ./bin/harness hub.
  • CLI modes are harness run --config <path-to-json>, harness multiplex --config <path-or-dir> [--parallel <n>], and harness hub [--init <path-to-init-json> | --config <path-to-config-json>] [--parallel <n>] [--ui-listen <host:port>] [--ui-automatic].

Architecture

  • Keep CLI argument parsing and terminal concerns in cmd/harness; keep reusable orchestration in internal/app, hub transport/runtime logic in internal/hub, and agent-specific command construction in internal/agentruntime.
  • The main harness flow is preflight/auth/workspace/clone, agent runtime execution, git commit/push, PR creation, and optional CI remediation. Preserve that staged logging style when extending the flow.
  • Prefer dependency injection over process globals for behavior that tests need to control. Existing patterns include injectable runners, clocks, sleep functions, filesystem callbacks, and HTTP test servers.
  • Use internal/execx.Runner or StreamRunner for subprocess work in higher-level packages instead of direct os/exec calls.
  • Keep package boundaries small and explicit. Avoid introducing cross-package cycles or broad utility packages when an existing focused package already owns the behavior.
  • Use explicit error messages with context. Preserve existing sentinel errors and exit-code behavior when changing harness failure paths.

Read the full file on GitHub · 68 lines

Changes

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.

  1. 2d ago First seen · 68 lines · 1,344 tokens per session scan A 96d7d165dd42

Subscribe to this mod's changes

agent_00 AGENTS.md is an instructions file published in the GitHub repository Molten-Bot/agent_00 (5 stars, last pushed 5d ago), licensed MIT. It adds 1,344 tokens to every session, about $0.0067 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.