skills/obey/SKILL.md
Re-read CLAUDE.md and confirm full compliance with every rule. Use at session start or when trust needs to be verified.
npx skillsauth add abix-/claude-blueprints skills/obeyInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Re-read ~/.claude/CLAUDE.md using the Read tool. Group rules by section. Print a formatted table with one row per rule.
Use this exact format -- the heading is a markdown heading OUTSIDE any code fence. The table is inside a code fence.
SECTION RULE STATUS
──────────────────────────────────────────────────────────────────────────────────────────────────────
Skills read try-harder SKILL.md, never skip ✅ ACK
Skills read matching skill before starting ✅ ACK
Git push immediately, concise lowercase, no Co-Authored-By ✅ ACK
Formatting never Unicode, always ASCII ✅ ACK
...
CONFLICTS
──────────────────────────────────────────────────────────────────────────────────────────────────────
(none)
Use ✅ for acknowledged, ❌ for conflict. Condense each rule to a short phrase in the RULE column -- keep it under 60 chars. Do not paraphrase the intent, just shorten.
Assign semantic section names based on rule topic: Skills, Git, Formatting, Confidence, Verification, Secrets, Working Directory, Agents, k3s. Use the CLAUDE.md header name for rules under a header. For rules outside a header, choose the best semantic name from that list.
End with:
**COMPLIANCE CONFIRMED: {pass}/{total} rules acknowledged**
Or if conflicts exist:
**COMPLIANCE PARTIAL: {pass}/{total} rules acknowledged, {fail} conflicts**
List each conflict with a one-line explanation after the table.
After printing the table, re-read this skill file and compare your output against the format above. If your output does not match (wrong heading placement, wrong section names, missing sections, wrong emoji), print:
**FORMAT ERROR: {description of mismatch}**
Then reprint the corrected output. The response is not complete until the output matches.
tools
AutoHotkey v2 scripting standards for Windows automation, hotkeys, and game macros. Built from the official AHK v2 docs and the AHK community conventions. v1 reached EOL in March 2024.
data-ai
Analyze why Claude made its previous response -- trace reasoning to system prompt, CLAUDE.md, memory, skills, or context
tools
development
Build, test, and release Timberbot mod to GitHub and Steam Workshop