agentwork-exchange AGENTS.md

Repository instructions for AgentWork Exchange, a project that makes personal AI agents discoverable, testable, reviewable, and suitable for defined jobs. They also describe local commands and contribution rules.

In plain words
What is it for?
Use them when setting up the project, adding agent-framework detectors, changing agent schemas, creating certification fixtures, or improving connector code.
Why use it?
They reduce the risk of committing generated files or private data and help keep connectors local, consent-based, and compatible with the project’s public interfaces.

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/earthwalking/agentwork-exchange/agents-md
Clone the repo
git clone --depth 1 https://github.com/earthwalking/agentwork-exchange

Made for: Codex, OpenCode.

Per session 330 This file is loaded in full into every session.
When invoked 330 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 $0.00330 $0.00330
Opus 5 $0.00165 $0.00165
Sonnet 5 $0.00066 $0.00066
Haiku 4.5 $0.00033 $0.00033

Measured 3d ago against content hash a8f770af15f2, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agentwork-exchange 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 3d 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 · 50 lines

What it actually says

Agent Contributor Guide

AgentWork Exchange is designed to be worked on by human maintainers and coding agents. If you are an agent reading this repository, start here.

Mission

Help personal agents become portable workforce units:

  • discoverable through local connectors
  • declared through agentwork.yaml
  • certified through reproducible tasks
  • matched to enterprise Job Specs
  • reviewed by humans for risky actions
  • tracked through auditable delivery and settlement records

Local Commands

pnpm install
pnpm build
pnpm dev
pnpm agentwork init --output .\agentwork.yaml --force
pnpm agentwork connect --file .\examples\agentwork.yaml

Contribution Rules

  • Do not commit generated manifests, logs, dist/, or node_modules/.
  • Do not add real credentials, real private paths, or customer data.
  • Keep connector behavior consent-first and local-first.
  • Prefer deterministic mock services unless a real integration is explicitly scoped.
  • Update README, schemas, and examples when changing public interfaces.

Good Agent Tasks

  • Add a detector for a new local agent framework.
  • Improve the Agent Passport schema.
  • Add deterministic certification fixtures.
  • Improve Bounty matching and scoring.
  • Add tests around connector import validation.
  • Make the UI clearer for agent owners and enterprise operators.

Safety Boundary

The MVP is not a production credential runner. Any feature that reads private files, starts external jobs, writes to enterprise systems, pays money, or posts to a third-party platform must keep a human review step and an audit event.

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. 3d ago First seen · 50 lines · 330 tokens per session scan A a8f770af15f2

Subscribe to this mod's changes

agentwork-exchange AGENTS.md is an instructions file published in the GitHub repository earthwalking/agentwork-exchange (5 stars, last pushed 2mo ago), licensed MIT. It adds 330 tokens to every session, about $0.0016 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.