Context Guardian

Prevent and manage context rot — the gradual decline in AI output quality during long conversations. Use this skill ALWAYS — it runs passively in every sessi...

This skill appears coherent and does what it says: it will read today's and yesterday's note files and write summaries to memory/YYYY-MM-DD.md (and optionally MEMORY.md/AGENTS.md). Before installing, consider: 1) Where the memory/ folder will live — conversation notes may contain sensitive info, so put memory/ outside any repo you publish (add to .gitignore) or use an encrypted workspace. 2) Review existing memory/ and MEMORY.md files — the agent will read them at session start. 3) Decide whether you want automatic persistent saves; if not, tell the agent to ask before saving or disable the skill for passive runs. 4) Note the minor metadata mismatch: the docs say “Use this skill ALWAYS” but the registry flag is always:false — enable/disable behavior is up to you. Overall, the skill is internally consistent but will create persistent files you should manage and protect.

Type: OpenClaw Skill Name: normieclaw-context-guardian Version: 1.0.1 The skill is a utility designed to help AI agents manage long-term conversation context by persisting summaries, decisions, and preferences to a local 'memory/' directory. It provides structured instructions for tracking session health, handling topic shifts, and performing periodic memory pruning without any evidence of data exfiltration, malicious execution, or unauthorized access to sensitive system files (SKILL.md, references/setup-guide.md).