0 Fork 0

mirror of https://github.com/kevinveenbirkenbach/computer-playbook.git synced 2026-11-09 20:26:33 +00:00

Files

T

kevinveenbirkenbachandClaude Opus 4.6 7753d3b7be docs: add interaction rule for permitted commands and code execution section

Prefer permitted host commands over approval-required ones, and run
non-permitted commands inside the container at /opt/src/infinito via
make exec to avoid permission prompts.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

2026-03-29 01:12:41 +01:00

1.7 KiB

Raw Blame History

CLAUDE.md

Startup — MUST DO at the Start of Every Conversation

You MUST read AGENTS.md and follow all instructions in it at the start of every conversation before doing anything else. Do NOT skip this, even for short or simple requests.

Interaction Rules

Questions MUST NOT lead to modifications, manipulation of files, code, or state.
Only explicit commands MAY trigger modifications or manipulation.
You MUST prefer commands that are already permitted in .claude/settings.json over commands that require interactive approval. If an equivalent permitted command exists, you MUST use it instead of the restricted one.

Code Execution

You SHOULD run permitted commands (listed in .claude/settings.json) directly on the host.
For commands that are NOT permitted on the host, you MUST run them inside the application containers instead. Use make up (or the appropriate Make target) to start the stack, then use make exec to open a shell inside the container.
The repository is mounted into the container at /opt/src/infinito (see compose.yml), so code changes are immediately available there.
This avoids permission prompts and keeps the workflow uninterrupted.

Configuration

Project-level permissions and sandbox rules are defined in .claude/settings.json.
See code.claude.com for documentation on how to modify it.

Documentation

See the Claude Code documentation for further information. For human contributor guidance on working with agents, see here.

1.7 KiB Raw Blame History