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/kagan-sh/kagan/agents-mdgit clone --depth 1 https://github.com/kagan-sh/kaganWrote 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.
[](https://agentmods.dev/instructions/kagan-sh/kagan/agents-md)<a href="https://agentmods.dev/instructions/kagan-sh/kagan/agents-md"><img src="https://agentmods.dev/badge/instructions/kagan-sh/kagan/agents-md.svg" alt="Measured on agentmods" height="20"></a>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 | $0.02051 | $0.02051 |
| Opus 5 | $0.01026 | $0.01026 |
| Sonnet 5 | $0.00410 | $0.00410 |
| Haiku 4.5 | $0.00205 | $0.00205 |
Grade A, and why
kagan 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 5d 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 — 118 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Kagan is an OpenCode plugin that supervises coding tasks on a kanban board, with one git worktree
per task. Before changing lifecycle behavior, read .specs/README.md for the
spec authority and read order. src/domain/task/metadata.ts is the authoritative metadata schema.
Commands
- Use Bun 1.3 or newer; CI pins 1.3.14. Run
bun install, thenbun run setuponce after cloning to enable the.githooks/pre-commithook. bun run verifyis the agent-facing curated gate. It runs the declared source complexity, comment, and circular-dependency checks, auto-formats with oxfmt, plus the localtestscript.bun run checkis the full check-only gate used by pre-commit and CI. It runsverifyx all --check, including its automatic test step, then package validation. CI checkout fetches full history so the new-comment check compares against the PR merge base.- Run one built-in check with
bunx verifyx lint,bunx verifyx format,bunx verifyx check-types, orbunx verifyx duplicate-code. Usebun run packagefor package checks. verify:complexityis two-tier: pure logic (src/**/*.tsexcludingsrc/tui/**andsrc/tui.tsx) must clear maintainability index 52; the TUI surface (src/tui/**/*.{ts,tsx},src/tui.tsx) clears 50 because its JSX render functions are inherently lower-scoring. Simplify the failing function first; split only when responsibilities genuinely diverge. Merging or unsplitting is valid when cohesion beats MI churn from artificial file boundaries. Useverifyx complexity --sloc-scope functionto score SLOC from each function body instead of the whole file. Never game the metric by deleting comments, joining lines, or fragmenting a coherent function. The exact command is pinned bytest/guards/validation.test.ts.- Run the full suite with
bun run test, not barebun test. The script supplies--conditions browser;bunfig.tomlsupplies the Solid preload. Bun positional test filters can also match a local gitignoredreferences/checkout because its exclude applies only to test discovery. - Focus a test with an anchored path and the browser condition, for example
bun test ./test/domain/task/policy.test.ts --conditions browser. bun run plugin:installinstalls a raw development snapshot globally. Re-run it after source edits and restart OpenCode.bun run plugin:install:prodinstalls the packed production artifact;bun run plugin:resetremoves Kagan from global OpenCode config.
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.
- 5d ago First seen · 118 lines · 2,051 tokens per session scan A ec09ce253dde
kagan AGENTS.md is an instructions file published in the GitHub repository kagan-sh/kagan (10 stars, last pushed 1mo ago), licensed MIT. It adds 2,051 tokens to every session, about $0.0103 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
collattice CLAUDE.md
Claude Code instructions for MrBildo/collattice, covering collattice, repository rules — hard, tech stack, build & run and prerequisites.
sanban AGENTS.md
AGENTS.md instructions for aancw/sanban, covering agents.md — guidance for ai agents, project, file map, api base and common tasks.
claudine AGENTS.md
Instructions for salam/claudine, covering atomic commits, specs, issue and bug reporting and fixing, implementation and misc.
claudine CLAUDE.md
Instructions for salam/claudine, a project described as: A kanban board for managing Claude Code conversations.
kanban CLAUDE.md
Claude Code instructions for kanban-rs/kanban, covering claude.md, project overview, architecture philosophy, solid principles applied and workspace structure.
flux AGENTS.md
Instructions for sirsjg/flux, covering agents.md, project overview, dogfooding, common commands and development.