What is jan-subagent AGENTS.md?

A set of repository instructions stored in an AGENTS.md file, telling coding agents how to work in a specific project, including its folder, structure, and required reading order.

Use it to define a repository summary, working directory, startup reading list, important modules, and engineering decisions for new coding sessions.

It gives an agent the project context it needs before making changes. This reduces mistakes caused by using the wrong files, commands, or assumptions.

jan-subagent AGENTS.md — Instructions file for Codex and OpenCode by pdurlej: A set of repository instructions stored in an AGENTS.md…

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/pdurlej/jan-subagent/agents-md

Clone the repo

git clone --depth 1 https://github.com/pdurlej/jan-subagent

Made for: Codex, OpenCode.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/pdurlej/jan-subagent/agents-md.svg)](https://agentmods.dev/instructions/pdurlej/jan-subagent/agents-md)

Your own site

<a href="https://agentmods.dev/instructions/pdurlej/jan-subagent/agents-md"><img src="https://agentmods.dev/badge/instructions/pdurlej/jan-subagent/agents-md.svg" alt="Measured on agentmods" height="20"></a>

Per session 932 This file is loaded in full into every session.

When invoked 932 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.

Model	Per session	Once invoked
Fable 5	$0.00932	$0.00932
Opus 5	$0.00466	$0.00466
Sonnet 5	$0.00186	$0.00186
Haiku 4.5	$0.00093	$0.00093

Measured 3d ago against content hash a998ecfe941d, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

jan-subagent 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 · 75 lines

How it starts

The opening of the file, as written. The whole thing — 75 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Scope

This file is local to /Users/pd/Developer/jan and should be treated as the primary Codex context for new threads started in this repository.

Working Folder

Always work from /Users/pd/Developer/jan.
Treat this directory as the repo root.

Repo Summary

Project: jan-subagent
Purpose: repo-native MCP agent for Polish-language change communication in software delivery, powered by NVIDIA Bielik.
Canonical runtime entrypoint: jan.jan_subagent_opencode
Legacy module jan/jan_subagent.py was removed in 2.0.0.
Current behavioral release target: 3.0.0.

Read Order For New Threads

/Users/pd/Developer/jan/docs/new-thread-start.md
/Users/pd/Developer/jan/README.md
/Users/pd/Developer/jan/docs/VERSION_3_0_0.md
/Users/pd/Developer/jan/_bmad-output/project-context.md
/Users/pd/Developer/jan/jan/jan_subagent_opencode.py
/Users/pd/Developer/jan/jan/workflow_engine.py
/Users/pd/Developer/jan/jan/policy.py
/Users/pd/Developer/jan/jan/source_adapters.py
/Users/pd/Developer/jan/tests/test_repo_native_workflows.py

Current Engineering Decisions

The opencode server is the default implementation for MCP usage and packaging.
setup.py version is aligned to 3.0.0.
jan/__init__.py uses lazy import for mcp.
- Do not reintroduce eager import of jan_subagent_opencode there.
Config is loaded from environment or ~/.jan/config.json.
Default model ID is speakleash/bielik-11b-v2.6-instruct.
README examples and API docs should stay aligned with named repo-native workflows and jan.workflow_engine.
mcp_config.json is the only maintained sample config and must stay valid JSON.
Primary workflows are:
- write_pr_description
- compose_release_notes
- rewrite_issue
- write_rollout_note
Legacy tools remain callable for compatibility, but are deprecated and not the main product surface.
jan.yml is the repo-level policy pack and should be treated as memory-as-code.
response_mode="review" must stay stable and JSON-only.
BMAD is present in two layers:
- maintenance scaffold: scripts/, references/, agents/, state/
- project workflows: _bmad/ and _bmad-output/
project-context.md in _bmad-output/ is the source-of-truth for the v3.0.0 pivot.
Benchmarking now has two tiers:
- workplace writing pilot as regression guardrail
- delivery benchmark v2 as north-star workflow benchmark

Read the full file on GitHub · 75 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.

3d ago First seen · 75 lines · 932 tokens per session scan A a998ecfe941d

Subscribe to this mod's changes

jan-subagent AGENTS.md is an instructions file published in the GitHub repository pdurlej/jan-subagent (0 stars, last pushed 5mo ago), licensed MIT. It adds 932 tokens to every session, about $0.0047 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.

NVIDIA bielik jan-kochanowski language-correction mcp polish-language

NemoClaw AGENTS.md

AGENTS.md instructions for NVIDIA/NemoClaw, covering agent instructions, project overview, product scope gate, agent skills and architecture.

NVIDIA/NemoClaw · 4,214 tokens

OpenNOW copilot-instructions.md

Copilot instructions for OpenCloudGaming/OpenNOW, covering copilot instructions for opennow, project scope, build, check, and packaging commands, from repository root and from opennow-stable/.

OpenCloudGaming/OpenNOW · 683 tokens

OpenNOW AGENTS.md

AGENTS.md instructions for OpenCloudGaming/OpenNOW, covering agents.md, core priorities, repository layout, platform layout (multi-provider ready) and module boundaries.

OpenCloudGaming/OpenNOW · 1,528 tokens

nvcf AGENTS.md

Instructions for NVIDIA/nvcf, covering agents.md - guide for ai coding agents, repo layout, cross-repo and stack routing, oss snapshot hygiene and local qa and testing environment safety.

NVIDIA/nvcf · 5,219 tokens

homelab-monitor CLAUDE.md

Claude Code instructions for SikamikanikoBG/homelab-monitor, covering claude.md and conventions.

SikamikanikoBG/homelab-monitor · 138 tokens

KimodoUnityBridge AGENTS.md

AGENTS.md instructions for OneYoungMean/KimodoUnityBridge: This root file is the repository-tool entry point. Detailed maintenance rules are in development/AGENTS.md; animation execution rules are in SKILL.md, and command details are in Command/help.json.

OneYoungMean/KimodoUnityBridge · 82 tokens