agent-configs/README.md

# Agent Configurations

Reusable working instructions for AI coding agents. Merge them into your own setup instead of replacing it.

## Configuration Index

| Configuration | Purpose |
| --- | --- |
| [Claude Sonnet 5.5 instructions](./claude-code/claude-sonnet-5-5/) | Latest model-specific `CLAUDE.md`: the Opus 5.5 file plus three short additions for behavior Anthropic's Sonnet 5.5 guide describes, and effort notes. |
| [Claude Opus 5.5 instructions](./claude-code/claude-opus-5-5/) | Model-specific `CLAUDE.md`: the Opus 5 conventions and Context7 guidance with fewer repeated or rigid rules, plus effort-level notes. |
| [Claude Opus 5 instructions](./claude-code/claude-opus-5/) | Model-specific `CLAUDE.md` with working conventions and Context7 guidance. |

## Where the file goes

`CLAUDE.md` is context, not an enforced policy. Place it where Claude Code loads it, for yourself across projects or for one project, and check which files loaded with `/memory`. See [Claude Code memory](https://code.claude.com/docs/en/memory). Anything that must be enforced belongs in [permissions](https://code.claude.com/docs/en/permissions) or [hooks](https://code.claude.com/docs/en/hooks-guide).

## Related Post

[CLAUDE.md for Opus 5 based on Anthropic's official docs](https://www.reddit.com/r/ClaudeAI/comments/1vd57c0/claudemd_for_opus_5_based_on_anthropics_official/)
Re-organize repo into prompt-templates and agent-configs 2026-08-02 03:27:54 +02:00			`# Agent Configurations`

Add the AI App Starter and reorganize the toolkit into three parts 2026-09-24 19:47:35 +02:00			`Reusable working instructions for AI coding agents. Merge them into your own setup instead of replacing it.`
Re-organize repo into prompt-templates and agent-configs 2026-08-02 03:27:54 +02:00
			`## Configuration Index`

Added a comprehensive Claude Code permission guard section 2026-09-07 18:29:45 +02:00			`\| Configuration \| Purpose \|`
			`\| --- \| --- \|`
Give each planning step one name and make the guides easier to follow 2026-10-12 14:57:27 +02:00			\| [Claude Sonnet 5.5 instructions](./claude-code/claude-sonnet-5-5/) \| Latest model-specific `CLAUDE.md`: the Opus 5.5 file plus three short additions for behavior Anthropic's Sonnet 5.5 guide describes, and effort notes. \|
			\| [Claude Opus 5.5 instructions](./claude-code/claude-opus-5-5/) \| Model-specific `CLAUDE.md`: the Opus 5 conventions and Context7 guidance with fewer repeated or rigid rules, plus effort-level notes. \|
			\| [Claude Opus 5 instructions](./claude-code/claude-opus-5/) \| Model-specific `CLAUDE.md` with working conventions and Context7 guidance. \|
Re-organize repo into prompt-templates and agent-configs 2026-08-02 03:27:54 +02:00
Add the AI App Starter and reorganize the toolkit into three parts 2026-09-24 19:47:35 +02:00			`## Where the file goes`
Re-organize repo into prompt-templates and agent-configs 2026-08-02 03:27:54 +02:00
Add the AI App Starter and reorganize the toolkit into three parts 2026-09-24 19:47:35 +02:00			`CLAUDE.md` is context, not an enforced policy. Place it where Claude Code loads it, for yourself across projects or for one project, and check which files loaded with `/memory`. See [Claude Code memory](https://code.claude.com/docs/en/memory). Anything that must be enforced belongs in [permissions](https://code.claude.com/docs/en/permissions) or [hooks](https://code.claude.com/docs/en/hooks-guide).
Update README with related post section 2026-09-07 17:38:06 +02:00
			`## Related Post`

			`[CLAUDE.md for Opus 5 based on Anthropic's official docs](https://www.reddit.com/r/ClaudeAI/comments/1vd57c0/claudemd_for_opus_5_based_on_anthropics_official/)`