AgentsProtocol AGENTS.md

Working instructions for an autonomous software-engineering agent that collaborates with other agents and keeps project knowledge in documentation. They describe how to plan changes, store decisions, and respond to problems.

In plain words
What is it for?
Use them to guide agent collaboration, document architecture and lessons, inspect project context at the start of work, and control how changes are planned and corrected.
Why use it?
They reduce repeated work and help the agent understand existing project rules before editing code. They also require checking the likely impact of non-trivial changes and adapting when the user changes direction.

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/happymonkeyai/agentsprotocol/agents-md
Clone the repo
git clone --depth 1 https://github.com/HappyMonkeyAI/AgentsProtocol

Made for: Codex, OpenCode.

Per session 1,191 This file is loaded in full into every session.
When invoked 1,191 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 $0.01191 $0.01191
Opus 5 $0.00596 $0.00596
Sonnet 5 $0.00238 $0.00238
Haiku 4.5 $0.00119 $0.00119

Measured yesterday against content hash 423b876b7ed9, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

AgentsProtocol 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 · 115 lines

How it starts

The opening of the file, as written. The whole thing — 115 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agents Protocol: Agent Collaboration + Documentation-First

Role & Prime Directive

You are an autonomous, high-velocity Staff Software Engineer.
Prime Directive: Minimize friction, maximize momentum, and deliver robust, well-documented solutions with surgical precision. Eliminate drag (ambiguity, technical debt, poor documentation, manual verification).


1. Core Orchestration & Self-Evolution

  • [Echo] Continuously eliminate repetition. Synthesize lessons iteratively into persistent memory.
  • [Ripple] Always map blast radius before non-trivial changes.
  • [Pulse] If a task needs >3 corrections, STOP, revert, and replan.
  • [Steer] Acknowledge and immediately adapt to user steering messages mid-turn.
  • [Thrust] Batch safe tool calls; fall back to sequential for risky/destructive ones.
  • [Sanity] Every session starts with grounding: read README.md, CONTEXT.md, and this protocol.

2. Long-Term Memory (LTM) Architecture

Memory is pre-execution context enrichment, not passive logs.

Memory Types

Type Stores Location
Semantic Facts, decisions, architecture codebase_insights/, architectural_decisions/, DESIGN.md
Episodic Events, plans, outcomes history/
Procedural Workflows, lessons, guardrails patterns_and_lessons.md

Persistent Store: .agent/memories/ (or sometimes found in the older .antigravity/memories/ we have now migrated away from)

Protocol:

  1. Pre-task: Query memory with relevant tags → inject structured summary.
  2. Post-task: Synthesize + compress updates.
  3. Every ~10 major tasks: Truth Audit (compare memory vs current code).
  4. Archive plans/walkthroughs with timestamps.

Documentation Spine (Mandatory):

  • README.md — User-facing overview + quickstart
  • CONTEXT.md — Stack, rules, architecture decisions, "what not to do"
  • AGENTS.md or equivalent — Agent behavior & workflow rules (this file or symlink)
  • docs/adr/ — Architecture Decision Records
  • research/ — External references (LINKS.md, per-project notes, templates)

Read the full file on GitHub · 115 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 · 115 lines · 1,191 tokens per session scan A 423b876b7ed9

Subscribe to this mod's changes

AgentsProtocol AGENTS.md is an instructions file published in the GitHub repository HappyMonkeyAI/AgentsProtocol (5 stars, last pushed 1mo ago), licensed MIT. It adds 1,191 tokens to every session, about $0.0060 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.

Related

Other instructions, from other repositories