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/systemfsoftware/constitution/agents-mdgit clone --depth 1 https://github.com/systemfsoftware/constitutionWhat 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 | $0.01307 | $0.01307 |
| Opus 5 | $0.00654 | $0.00654 |
| Sonnet 5 | $0.00261 | $0.00261 |
| Haiku 4.5 | $0.00131 | $0.00131 |
Grade A, and why
constitution 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 yesterday.
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 — 122 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Constitution Repository
Single source of truth for the supreme design law of System F Software. Consumer repos vendor via git subtree + symlink. This repo has no production code, no test suite, and no build step — it is two markdown documents plus their governance tooling (commit validation, agent harness): CONSTITUTION.md, resident in every agent's context, and CONSTITUTION-ARTICLES.md, retrieved on write or edit of a source file.
@CONSTITUTION.md
Startup Workflow
Before making changes:
- Read this file completely.
- Confirm the active task with the user or the agent's task list.
- Review recent commits with
git log --oneline -5. - Ensure current branch is not
main— feature branches only. If on main, create one.
Working Rules
- One task at a time. Finish before starting the next.
- Conventional commits required. The commit-msg hook enforces
type(scope): description. Rungit committhrough the hook — do not bypass with--no-verify. - Verification required. Run the verification commands before claiming done.
- Stay in scope. Don't modify files unrelated to the task. Scope reduction requires explicit user approval.
- Leave clean state. The next session must run verification immediately.
Amending the Constitution
File Split
CONSTITUTION.md(Resident): Conduct rules (Article V + Governance) loaded in every session.CONSTITUTION-ARTICLES.md(Retrieved): Domain, boundary, verification, and organization rules (Articles I–IV) loaded on write/edit of source code.
Writing a Rule
Rules are fenced YAML blocks with: id, title, gate, do, dont, harm, check (and optional example, scope, layers).
Minting an ID
ID format: CONST-<family><n>. Pick the next free number in the family (never renumber to close gaps).
| Letter | Family | Purpose |
|---|---|---|
G |
Governance | Invoking constitution or resolving priority |
E |
Enforcement | Gate design, execution, verification |
P |
Purity | Decision functions and side-effect isolation |
D |
Domain modelling | Domain types and constraints |
B |
Boundary | Core/shell boundaries, effects, adapters |
T |
Testing | Testing strategy, mutation, properties |
N |
Naming & structure | Module organization and naming |
W |
Work discipline | Task scope, bypass declarations, reviews |
S |
Subtraction | Code deletion and structural simplification |
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.
- yesterday First seen · 122 lines · 1,307 tokens per session scan A 44a302f3105b
constitution AGENTS.md is an instructions file published in the GitHub repository systemfsoftware/constitution (2 stars, last pushed 2d ago), licensed Apache-2.0. It adds 1,307 tokens to every session, about $0.0065 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
open-agent-ready-typescript AGENTS.md
Instructions for lucasgodt/open-agent-ready-typescript, covering agents.md, what this repository is, commands, architecture map and the workflow: spec first.
open-agent-ready-typescript CLAUDE.md
Instructions for lucasgodt/open-agent-ready-typescript: Read AGENTS.md. It is the single source of agent instructions for this repo.
ai-project-template AGENTS.md
Instructions for TheSmuks/ai-project-template, covering project context, project overview, build & run, install dependencies and build.
vibeproof AGENTS.md
Instructions for HumanDealer/vibeproof, covering agents.md — operating contract for ai coding agents, the non-negotiables (memorize these), the standard workflow (every change), before you say "done" and when something breaks.
vibeproof CLAUDE.md
Instructions for HumanDealer/vibeproof, covering claude.md — claude code operating contract, the non-negotiables, claude code specifics and the standard workflow.
android-finance-spec AGENTS.md
Instructions for brycegao/android-finance-spec: When generating, editing, reviewing, or refactoring this repository, follow all Android finance specs.