prd-taskmaster AGENTS.md

Instructions for Atlas, a system in which a leading AI model breaks work into tasks and sends them to suitable command-line agents or direct APIs. The guide defines Atlas's goal and the rules used to judge changes.

In plain words
What is it for?
Use it when changing the driver, task decomposition, executor selection, validation checks, or other parts of Atlas's agent workflow.
Why use it?
It helps coding agents keep changes aligned with Atlas's task-routing design and verify that work is genuinely complete rather than merely appearing finished.

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/anombyte93/prd-taskmaster/agents-md
Clone the repo
git clone --depth 1 https://github.com/anombyte93/prd-taskmaster

Made for: Codex, OpenCode.

Per session 732 This file is loaded in full into every session.
When invoked 732 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 $0.00732 $0.00732
Opus 5 $0.00366 $0.00366
Sonnet 5 $0.00146 $0.00146
Haiku 4.5 $0.00073 $0.00073

Measured yesterday against content hash f6a65c20a07c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

prd-taskmaster 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 yesterday.

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 · 53 lines

How it starts

The opening of the file, as written. The whole thing — 53 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent Instructions — Atlas

Applies to ANY agent harness working in this repo (Codex, Gemini, Claude Code, or the next one). Claude Code also reads CLAUDE.md, which carries the same contract.

The end goal (north star — test every change against this)

Atlas's entry point is a frontier model acting as the driver. The driver understands the user's intent using skills built for it, then — inside a loop — decomposes the goal into optimal components and routes each to the cheapest capable executor, saving as many tokens as possible while delivering an outcome as good as if the frontier model had done all the work itself, at a fraction of the cost.

The executor pool is any CLI agent harness plus direct API paths, all in parallel — never a fixed provider list. Every feature decision is tested against: does this make the frontier-driver-decomposes-to-cheap-executors loop better, cheaper, or more provable?

The four moats (why this is a product, not a prompt pack):

  1. Cross-vendor arbitrage — route work across any harness/API by cost and capability.
  2. Engine-enforced unfakable gates — validate-tasks, ship-check (SHIP_CHECK_OK), structural tests. A feature isn't a product feature until the engine enforces it.
  3. Persistent vendor-neutral task graph — tasks.json survives across vendors/sessions.
  4. Cross-vendor cost ledger — telemetry + economy-report prove the savings.

Backend model (v4.1+)

TaskMaster is one optional pluggable backend. Backend resolution is backend: auto|taskmaster|native in .atlas-ai/fleet.json; the 5-op protocol (detect / init_project / parse_prd / expand / rate) lives in prd_taskmaster/backend.py.

  • Task-state ops are engine-native ALWAYS: python3 script.py next-task and python3 script.py set-status <id> <status> (MCP: next_task / set_task_status). Never call bare task-master outside a Mode-B-labeled block — structural tests enforce this.
  • No API keys → the backend returns a structured agent_action_required payload and YOU (the driving agent) do the generation, validated through validate-tasks before any write. Never a crash.
  • The "## Backend operations" table in SKILL.md is normative. When docs and code disagree, fix the docs to match the table.

Read the full file on GitHub · 53 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. yesterday First seen · 53 lines · 732 tokens per session scan A f6a65c20a07c

Subscribe to this mod's changes

prd-taskmaster AGENTS.md is an instructions file published in the GitHub repository anombyte93/prd-taskmaster (592 stars, last pushed 18d ago), licensed MIT. It adds 732 tokens to every session, about $0.0037 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.