docs: add signpost llms.txt + consolidate agent instructions into AGENTS.md

.github/copilot-instructions.md

@@ -1,24 +1,9 @@
 # Copilot Coding Agent Instructions
 
-You are working on **PyAutoConf**, the configuration management library for the PyAuto ecosystem.
+You are working on **PyAutoConf**, the configuration, serialization and I/O
+foundation of the PyAuto ecosystem (package `autoconf`).
 
-## Key Rules
-
-- Run tests after every change: `python -m pytest test_autoconf/`
-- All files must use Unix line endings (LF, `\n`)
-- If changing public API (function signatures, class names, import paths), clearly document what changed in your PR description — all downstream PyAuto packages depend on this
-
-## Architecture
-
-- `autoconf/conf.py` — Core configuration system
-- `autoconf/dictable.py` — Dictionary/serialization support
-- `autoconf/fitsable.py` — FITS file handling
-- `autoconf/json_prior/` — JSON-based configuration priors
-- `autoconf/tools/` — Utility decorators and helpers
-- `test_autoconf/` — Test suite
-
-## Sandboxed runs
-
-```bash
-NUMBA_CACHE_DIR=/tmp/numba_cache MPLCONFIGDIR=/tmp/matplotlib python -m pytest test_autoconf/
-```
+The canonical, agent-agnostic instructions for this repo live in **`AGENTS.md`** at
+the repository root — build and test commands, architecture, the base-layer config
+rules, and conventions. Read it directly (Copilot does not process `@`-imports) and
+treat it as the single source of truth.

AGENTS.md

@@ -22,6 +22,15 @@ on autoconf, so any public-API change here ripples downstream.
 - No `docs/` / RTD site — the package source and `test_autoconf/` are the
   authoritative reference.
 
+## Architecture
+
+- `autoconf/conf.py` — layered config system (`Config` / `conf.instance`).
+- `autoconf/dictable.py` — dict / JSON serialization (`output_to_json` / `from_json`).
+- `autoconf/fitsable.py` — FITS I/O (`output_to_fits` / `ndarray_via_fits_from`).
+- `autoconf/json_prior/` — JSON-based priors.
+- `autoconf/tools/` — shared decorators and helpers.
+- `test_autoconf/` — test suite.
+
 ## Quick commands
 
 ```bash

llms.txt

@@ -0,0 +1,15 @@
+# PyAutoConf
+
+> PyAutoConf (package `autoconf`) is the foundational configuration, serialization and I/O layer of the PyAuto ecosystem — layered config with overrides, dict/JSON/FITS serialization, JSON priors, and shared utilities every other library inherits. This file is a signpost: it points you to the right resource by intent — see where it is used, or work on the library itself.
+
+## Where it's used
+
+- PyAutoConf has no standalone workspace or tutorial series; it underpins every other PyAuto library. You meet it in practice through the consumer libraries — [PyAutoFit](https://github.com/PyAutoLabs/PyAutoFit), [PyAutoGalaxy](https://github.com/PyAutoLabs/PyAutoGalaxy), [PyAutoLens](https://github.com/PyAutoLabs/PyAutoLens) — and their workspaces.
+
+## Work on it (contributors / coding agents)
+
+- [AGENTS.md](./AGENTS.md): build, tests, architecture, base-layer config rules, and conventions.
+
+## Ecosystem
+
+- The base layer of the stack: depends on nothing else in the ecosystem; PyAutoArray, PyAutoFit, PyAutoGalaxy and PyAutoLens are all built on it.