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.
npx agentmods add instructions/vladev0/agentboard/agents-mdgit clone --depth 1 https://github.com/Vladev0/agentboardWhat 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 | $0.04236 | $0.04236 |
| Opus 5 | $0.02118 | $0.02118 |
| Sonnet 5 | $0.00847 | $0.00847 |
| Haiku 4.5 | $0.00424 | $0.00424 |
Grade A, and why
agentboard 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.
How it starts
The opening of the file, as written. The whole thing — 290 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AgentBoard — conventions for agents
AgentBoard is a task tracker for humans and AI agents, built on plain markdown files. There are two ways to work with tasks: MCP tools (the primary path) and direct file access (a fallback, if MCP isn't available). Both paths write the same format, so files stay consistent no matter which one you use.
Quick start for a new agent
list_projects— if the list is empty, there's no project yet:create_project(name, key?), thencreate_taskfor the first tasks. If projects already exist, pick one byslug.get_next_task(project)— don't read the whole task list, just ask what to do.- For the task returned —
get_task(project, id)to see the full description, updates, comments, and subtasks before starting work. - Work (see the tool table below), logging updates/comments as you go.
- Planning to work in cycles/iterations without a human in the loop each time? See "Protocol for long loop/cycle sessions" below — it's written for exactly that case.
Three different kinds of entries on a task — don't confuse them
| Type | Lives in | When to use |
|---|---|---|
| Comment | ## Comments |
Discussion, a clarifying question, a passing remark while working. Not versioned. |
| Update (big) | ## Updates, versions vN |
A deliberate checkpoint: the task's substance (its description) was rewritten after work/discussion led somewhere. Requires a summary of "what and why." Versioned, shown in the UI as the task's main evolution history. |
| Activity | ## Activity |
Routine technical bookkeeping: status changes, priority/field changes. Generated automatically, not versioned, collapsed in the UI by default. |
Practical rule: if you're just replying or clarifying, that's a comment (add_comment). If
the task's actual substance changes as a result of the discussion, that's an update
(update_description with a required summary). Never log an "update" without an actual
description change — that's what comments are for.
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.
- 2d ago First seen · 290 lines · 4,236 tokens per session scan A 814b23a98603
agentboard AGENTS.md is an instructions file published in the GitHub repository Vladev0/agentboard (1 stars, last pushed 1mo ago), licensed MIT. It adds 4,236 tokens to every session, about $0.0212 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.
Other instructions, from other repositories
kanban-md AGENTS.md
Instructions for antopolskiy/kanban-md, covering guidelines, releasing, ci verification (required), release notes and new: feature name (by @contributor).
kanban-md CLAUDE.md
Instructions for antopolskiy/kanban-md, a project described as: File-based kanban board for autonomous agentic loop. CLI and TUI for multi-agent workflows. Skills included.
kanban-md GEMINI.md
Instructions for antopolskiy/kanban-md, a project described as: File-based kanban board for autonomous agentic loop. CLI and TUI for multi-agent workflows. Skills included.
Backlog.md AGENTS.md
AGENTS.md instructions for MrLesk/Backlog.md, covering read this first, simplicity-first implementation rules, ui descriptions, commands and development.
data-olympus AGENTS.md
Instructions for knaisoma/data-olympus, covering data olympus, project overview, architecture, development workflow and code standards.
gsd-task-manager CLAUDE.md
Instructions for vscarpenter/gsd-task-manager, covering claude.md, project overview, design context, core commands and development.