wtcraft AGENTS.md

Repository instructions for wtcraft. They define how work is routed, which files are authoritative, how model information is checked, and how releases are handled.

In plain words
What is it for?
Use them when modifying wtcraft or its templates, following worktree task instructions, checking role-model configuration, and creating version tags.
Why use it?
They help agents follow the project’s worktree rules and avoid changing only a copied template or using outdated model information.

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/zywkloo/wtcraft/agents-md
Clone the repo
git clone --depth 1 https://github.com/zywkloo/wtcraft

Made for: Codex, OpenCode.

Per session 257 This file is loaded in full into every session.
When invoked 257 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 $0.00257 $0.00257
Opus 5 $0.00129 $0.00129
Sonnet 5 $0.00051 $0.00051
Haiku 4.5 $0.00026 $0.00026

Measured 2d ago against content hash 1c1be97017a0, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

wtcraft 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 · 24 lines

What it actually says

wtcraft routing

If .worktree-task.md exists in the current worktree root, read .agent-harness/executor.md first and follow task Scope, Off-limits, and Verification.

repo structure (wtcraft dogfooding)

templates/ is the source of truth for files wtcraft init copies to user repos. The corresponding live files in this repo are wtcraft's own dogfooded versions:

templates/.agent-harness/ ↔ .agent-harness/ templates/.claude/commands/ ↔ .claude/commands/

When changing harness behavior: update both the template AND the live file.

model knowledge policy

Never rely on your own training knowledge for model names or IDs — it is always outdated. For model recommendations, read .agent-harness/role-models.yml. If the user states a model name, that overrides everything else.

release guardrails

  • Version tags must use v<semver> format (example: v0.3.8).
  • Version tags must be created from main only.
  • Never create or move version tags from feature/worktree branches.
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 · 24 lines · 257 tokens per session scan A 1c1be97017a0

Subscribe to this mod's changes

wtcraft AGENTS.md is an instructions file published in the GitHub repository zywkloo/wtcraft (3 stars, last pushed 12d ago), licensed Apache-2.0. It adds 257 tokens to every session, about $0.0013 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.