LastHumanCommit AGENTS.md

Repository instructions for an agent role router that first inspects the workspace, branch, and project location. They also define where temporary files and user-requested worktrees may be placed.

In plain words
What is it for?
Use them before repository work, especially when checking worktrees, identifying the main checkout, handling temporary files, or creating an explicitly requested worktree.
Why use it?
They prevent agents from working in the wrong checkout, creating unexpected branches, or storing project material in unsafe locations.

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/megamen32/lasthumancommit/agents-md
Clone the repo
git clone --depth 1 https://github.com/megamen32/LastHumanCommit

Made for: Codex, OpenCode.

Per session 2,522 This file is loaded in full into every session.
When invoked 2,522 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 $0.02522 $0.02522
Opus 5 $0.01261 $0.01261
Sonnet 5 $0.00504 $0.00504
Haiku 4.5 $0.00252 $0.00252

Measured yesterday against content hash 317d75b8e73c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

LastHumanCommit 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.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

AGENTS.md · 223 lines

How it starts

The opening of the file, as written. The whole thing — 223 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent role router

Workspace first

Before task work, inspect the repository root, git worktree list --porcelain, the current branch or detached HEAD, and the default branch when identifiable.

Routine work stays in the current primary checkout. Do not create, switch, merge, or delete a branch or worktree for isolation, cleanliness, review, or an ordinary task. If the harness started in an auxiliary worktree, detached HEAD, or a non-default branch, the first user-visible update must warn the user and show the exact worktree path, branch, and primary checkout.

If the user explicitly asks LHC to create a worktree, create it only at <primary-project-root>/.worktrees/<task-slug>. Never create a project worktree in /tmp, a home cache, a sibling directory, or harness-owned storage. If the harness already selected another checkout, do not create a second one or move silently. Follow src/common/protocols/SHARED_WORKTREE.md for concurrent edits.

Project-local temporary storage

All temporary project material lives under <project-root>/.tmp/, which the project must ignore in Git. This includes source code, repository clones and exports, build trees and caches, binaries, packages, APK/DMG files, archives, checksums, and release artifacts, even when they exist only briefly. Never put such material in system /tmp, $TMPDIR, or a language runtime's default temp directory. System temp is allowed only for tiny non-code OS primitives when an OS or API genuinely requires it, never for project data or deliverables.

Resolve one role

If an enclosing instruction explicitly assigns one of these roles, read only that role file and follow it:

  • Lead: src/common/agents/Lead.md
  • Overseer: src/common/agents/Overseer.md
  • Worker: src/common/agents/Worker.md
  • Tester: src/common/agents/Tester.md
  • Reviewer: src/common/agents/Reviewer.md

Do not read unrelated role prompts. If it says you are a subagent but assigns no known role, stop and ask L; never promote yourself to Lead. Otherwise you are L: read src/common/agents/Lead.md.

Read the full file on GitHub · 223 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 · 223 lines · 2,522 tokens per session scan A 317d75b8e73c

Subscribe to this mod's changes

LastHumanCommit AGENTS.md is an instructions file published in the GitHub repository megamen32/LastHumanCommit (2 stars, last pushed 2d ago), licensed MIT. It adds 2,522 tokens to every session, about $0.0126 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.