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/nrdxp/predicate/agents-mdgit clone --depth 1 https://github.com/nrdxp/predicateWhat 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 | $0.01551 | $0.01551 |
| Opus 5 | $0.00776 | $0.00776 |
| Sonnet 5 | $0.00310 | $0.00310 |
| Haiku 4.5 | $0.00155 | $0.00155 |
Grade A, and why
predicate 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 — 124 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Predicate
This repository is Predicate, the upstream. skills/, rules.md, and
ambient.md live at the root because other projects consume them as an installed
plugin — see README.md and
docs/getting-started.md.
Goal
Predicate is a self-hosting, harness-agnostic framework that keeps AI coding agents anchored to the true goal across long-horizon work. It has two synergistic halves:
- Correction — externalize correctness to the strongest evaluator (the Verification Dual: verify, then trust — no condition closed by an agent's say-so). Authority: rules.md.
- Prevention — externalize the goal, requirements, unknowns, and available
tools as a durable, selectively-projected conditioning layer (the always-on
conditioning/system-prompt law, the/orientworkflow, the nested AGENTS.md hierarchy), so a walk stays focused before drift compounds.
Correction without prevention corrects toward the wrong goal; prevention without correction drifts anyway over a long horizon. Together they bound drift.
Formal substrate
An LLM is an autoregressive stochastic walk over a token state-space; in open-loop generation error compounds, so drift is a statistical inevitability over long horizons. Predicate closes the loop — an external deterministic evaluator computes an error differential and updates the boundary condition toward a fixed point (ΔE → 0). The formalism is non-entropic: it is the anchor the doctrine rests on, not the doctrine. Full treatment: docs/theory/formalism.md; lexicon in rules.md §1. The prevention extension — attention-dilution and phase-space constriction — is derived in docs/theory/formalism.md Part 2.
Requirements (for predicate to be useful)
- Self-hosting — predicate governs its own development.
- Harness-agnostic — every capability degrades to a harness-native primitive
(
git/bash/python/nickel); a harness convenience may accelerate a step but never replaces the primitive path when it is absent or fails. - Machine-checkable where possible — artifacts and process are bound by evaluators (footprints), not by an agent's memory of the rules.
- Composable — co-exists with whatever other plugins / skills / MCP servers the host harness has installed.
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 · 124 lines · 1,551 tokens per session scan A ef03c74bac85
predicate AGENTS.md is an instructions file published in the GitHub repository nrdxp/predicate (10 stars, last pushed 8d ago), licensed MIT. It adds 1,551 tokens to every session, about $0.0078 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
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
buildNext
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
spec-kit AGENTS.md
Instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
langchain AGENTS.md
Instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.