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/navikt/copilot/agents-mdgit clone --depth 1 https://github.com/navikt/copilotWrote 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/navikt/copilot/agents-md)<a href="https://agentmods.dev/instructions/navikt/copilot/agents-md"><img src="https://agentmods.dev/badge/instructions/navikt/copilot/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 | $0.00720 | $0.00720 |
| Opus 5 | $0.00360 | $0.00360 |
| Sonnet 5 | $0.00144 | $0.00144 |
| Haiku 4.5 | $0.00072 | $0.00072 |
Grade A, and why
copilot 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 today.
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 — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md for navikt/copilot
Minimal guide for agents working in this repository.
What this repo is
Monorepo for Nav's Copilot ecosystem:
apps/my-copilot(Next.js/TypeScript web app)apps/copilot-api(Go backend API)apps/copilot-metrics(Go metrics job)apps/mcp-onboarding(Go MCP reference server)apps/mcp-registry(Go MCP registry API)
SECURITY.md has the security model and trust boundaries.
Efficiency rule
Keep tool output small. Run targeted commands, do not dump whole files or full build logs into context, and reach for deterministic tools (rg, git, gh) before asking the model to guess.
rtk is an optional CLI proxy that filters terminal output before it reaches the model. Use it if you want to (rtk git status). Nothing here requires or assumes it, and public controlled measurement has not reproduced its advertised savings (JetBrains study).
Standard commands
From repo root:
mise check
mise test
mise build
mise all
After editing an app, run mise check in that app's directory. Run mise all when the change reaches across the repo.
Conventions
- Start with the smallest safe change and keep diffs task-focused.
- Reuse existing patterns before adding new abstractions.
- In
my-copilot, use Aksel spacing tokens, not Tailwindp-*/m-*utilities. - Do not commit secrets.
- Do not push unless explicitly asked.
Running the nav-pilot binary in tests or verification
nav-pilot writes to the developer's own environment: ~/.copilot/,
~/.nav-pilot/ and the repository you are standing in. A verification that runs
the real binary without isolation changes a real setup, and it has: an install
run during testing pointed a user scope at a temporary worktree and left 23 of 52
files in conflict, which sync then silently skipped. Nothing had been edited by
hand.
Whenever you run the binary — directly, or through a test that calls run():
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.
- today Changed · +26 lines · +355 tokens per session e83d66d16de3
- 4d ago First seen · 43 lines · 365 tokens per session scan A 42f0b85b696d
copilot AGENTS.md is an instructions file published in the GitHub repository navikt/copilot (54 stars, last pushed today), licensed MIT. It adds 720 tokens to every session, about $0.0036 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 2026-08-30.
Other instructions, from other repositories
sensei AGENTS.md
Instructions for spboyer/sensei, covering agents.md, repository structure, key conventions, skill.md format and skill body (markdown).
sensei copilot-instructions.md
Instructions for spboyer/sensei, covering copilot instructions, project context, skill authoring rules, token and scoring expectations and validation commands.
SonarCloud-Skill-Agent-Guidance
Repository guidance for the SonarCloud/SonarQube findings management skill.
kasetto CLAUDE.md
Claude Code instructions for pivoshenko/kasetto, covering claude.md, what this repo is, commands, rust architecture and module map.
Copilot-Skills copilot-instructions.md
Copilot instructions for TheSethRose/Copilot-Skills, covering copilot instructions, find what you're looking for, get file context before editing, understand related code and orchestration default.
Copilot-Skills document-project.instructions.md
Auto-loaded guidance for generating project documentation across any codebase.