codex

A setup that gives Codex, a coding agent, lasting memory about a code repository. It teaches Codex when to retrieve that memory and automatically saves selected preferences and instructions after each conversation.

In plain words
What is it for?
It is for installing GPS in Codex projects through AGENTS.md, Codex configuration, a memory tool connection, and a turn-end hook. The setup lets Codex retrieve repository context and capture durable preferences automatically.
Why use it?
It reduces repeated explanations across coding sessions and helps Codex use project-specific decisions in later work. It also records instructions such as standing preferences or rules tied to parts of the codebase.

Agent

Install

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.

agentmods
npx agentmods add agents/invariance-ai/gps/codex
Clone the repo
git clone --depth 1 https://github.com/invariance-ai/gps
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 879 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.00879
Opus 5 $0.00000 $0.00439
Sonnet 5 $0.00000 $0.00176
Haiku 4.5 $0.00000 $0.00088

Measured 2d ago against content hash 20a492feebcb, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

codex 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.

docs/guide/agents/codex.md · 66 lines

How it starts

The opening of the file, as written. The whole thing — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Codex CLI

GPS gives Codex durable, symbol-anchored repo memory. Codex learns the workflow from AGENTS.md, gets structured tools through MCP, and uses the notify hook to capture durable preferences and directives after each turn.

npx -y @invariance/gps setup --yes --with-codex

Writes:

  • AGENTS.md — appends a <!-- gps:start -->...<!-- gps:end --> block. Codex reads this at session start and learns when to run GPS.
  • .codex/config.toml — appends a # gps:start ... # gps:end managed block with two entries:
    • notify = ["npx", "-y", "@invariance/gps", "attach", "--transcript", "-"] — turn-end hook that distills the transcript into Decisions and automatically captures preferences ("from now on…", "always…", "i prefer…") and location-scoped directives
    • [mcp_servers.gps] — registers gps serve as an MCP server

The notify hook is how Codex gets automatic preference and directive capture. It fires after every turn, reads the transcript via stdin, and persists durable instructions into the GPS memory layer according to your capture policy. You do not need to call record_preference or record_directive manually in Codex for normal "from now on" instructions; the hook handles that.

The Codex CLI does not expose pre-tool-use hooks, so the index is refreshed lazily by gps prepare and gps context calls. Treat gps like rg: a local CLI Codex runs before non-trivial edits.

Codex learns the workflow from AGENTS.md. The MCP block gives it tool access, and the notify hook captures what it should remember after each turn.

What Codex does with gps

The AGENTS.md block teaches Codex this loop:

gps prepare --intent "<what I am about to change>"
gps prepare <symbol> --intent "<short intent>"
gps brief
gps remember "<hard-won repo fact>"
gps remember "<unfinished follow-up>" --reminder --symbol <symbol>
gps lessons record "<…>"

When MCP is available, Codex can call prepare_edit, brief, get_context, tests_for, invariants_for, record_preference, and record_directive as tools. The MCP tools and CLI commands read the same .gps/ memory. Use whichever interface is more natural in the moment; the data is the same.

Read the full file on GitHub · 66 lines

Changes

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.

  1. 2d ago First seen · 66 lines · 0 tokens per session scan A 20a492feebcb

Subscribe to this mod's changes

codex is an agent published in the GitHub repository invariance-ai/gps (19 stars, last pushed 3mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 879 tokens. 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.