PhD-Zero AGENTS.md

A set of operating instructions for an AI research and development workspace, including rules for experiments, evaluations, debugging, and research tasks.

In plain words
What is it for?
Use it to guide research workflows, experiment execution, technical analysis, and other AI R&D tasks in the TenureAI or PhD-Zero workspace.
Why use it?
It gives the coding agent a consistent process for routing non-trivial work, checking its reasoning, and following workspace guardrails.

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/tenureai/phd-zero/agents-md
Clone the repo
git clone --depth 1 https://github.com/TenureAI/PhD-Zero

Made for: Codex, OpenCode.

Per session 3,031 This file is loaded in full into every session.
When invoked 3,031 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 $0.03031 $0.03031
Opus 5 $0.01515 $0.01515
Sonnet 5 $0.00606 $0.00606
Haiku 4.5 $0.00303 $0.00303

Measured yesterday against content hash 1d441ec93652, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

PhD-Zero 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 · 179 lines

How it starts

The opening of the file, as written. The whole thing — 179 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AI R&D Workspace

This workspace is for AI research and development tasks (reproduction, debugging, evaluation, training, and experiment planning).

⚠ Mandatory Skill Routing — MUST follow, violations are NOT acceptable

Every non-trivial task MUST go through the skill pipeline via the Skill tool. Do NOT answer research questions, run experiments, or write papers by generating freeform text without first invoking the corresponding skill.

Routing Order (execute top-down, stop at the first matching rule)

  1. research-workflow (PRIMARY ORCHESTRATOR) — invoke FIRST for any non-trivial task. All downstream skills below should be invoked FROM WITHIN research-workflow, not standalone.
  2. deep-research — MANDATORY when user message contains ANY research-intent keyword:
    • Chinese: 调研/研究/对比/综述/文献/证据/机制/根因/为什么/可行性/路线图/分析/探索
    • English: research/investigate/compare/survey/literature/evidence/mechanism/root-cause/why/feasibility/roadmap/analyze/explore
    • If ANY keyword matches → you MUST invoke deep-research. No exceptions. Skipping is a routing violation.
  3. experiment-execution — when user asks to run/launch/start/resume/monitor an experiment.
  4. research-plan — when user asks for a proposal, roadmap, ablation plan, or study design.
  5. paper-writing — ONLY when user explicitly asks to draft/write/revise a paper or section.
  6. project-context — when environment setup or runtime fields are needed before execution.
  7. run-governor — at run start to set mode + run_id.
  8. memory-manager — bootstrap at run start, writeback at task end, trigger-based in between.
  9. human-checkpoint — for safety risks, high-resource approvals, or hard blockers.

Self-Check Before Every Reply

Before producing any substantive response, you MUST run this mental checklist:

  1. Is this task non-trivial? → If yes, did I invoke research-workflow? If not, invoke it NOW.
  2. Does the user message contain any research-intent keyword from rule 2? → If yes, did I invoke deep-research? If not, invoke it NOW.
  3. Am I about to answer a research question with freeform text instead of skill output? → STOP. Invoke the skill first.

Read the full file on GitHub · 179 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 · 179 lines · 3,031 tokens per session scan A 1d441ec93652

Subscribe to this mod's changes

PhD-Zero AGENTS.md is an instructions file published in the GitHub repository TenureAI/PhD-Zero (52 stars, last pushed 5mo ago), licensed MIT. It adds 3,031 tokens to every session, about $0.0152 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.