context-scope-tags
Use when: you need strict context boundaries in chat (Telegram/Discord/Slack/etc.) and want to prevent topic bleed using explicit tags like [ISO], [SCOPE], [GLOBAL], [NOMEM], [REM]. Don’t use when: you want normal free-form conversation with automatic carry-over. Output: a copy/paste tag cheat sheet + routing rules.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/phenomenoner/context-scope-tagsContext Scope Tags (Chat Protocol)
A lightweight, portable convention for explicit context boundaries in chat.
Quick start
Put one or more tags at the very start of your message, then write normally.
Examples:
[ISO: FinLife] Debug the Windows run instructions. Don’t use any other repo context.[SCOPE: openclaw-mem] Implement the next benchmark step. Keep the answer scoped.[GLOBAL][REM] Remember: use UTC for cron schedules unless I say otherwise.[ISO: marketing][NOMEM] Draft 5 ad angles; do not store any long-term memory.
Tag parsing rules
- Tags must appear at the start of the user’s message.
- You may place multiple tags (recommended order: scope → memory intent).
- Tags do not override safety policies, tool policies, access controls, or approvals.
Supported tags
Isolation / scope
-
[ISO: <topic>]/[Isolated Context: <topic>]- Treat as a fresh topic.
- Do not pull in other conversation/project context unless the user explicitly re-provides it.
- Allowed implicit carry-over: universal safety rules + a few stable user prefs (timezone, “don’t apply config changes without approval”, etc.).
-
[SCOPE: <topic>]/[Scoped Context: <topic>]- Restrict reasoning to the named scope.
- If missing details inside the scope, ask clarifying questions.
-
[GLOBAL]/[Global Context OK]- Cross-topic reuse is allowed.
- When reusing prior context, call out what was reused.
Memory intent
-
[NOMEM]/[No Memory]- Do not store durable/long-term memories from this exchange.
-
[REM]/[Remember]- If the message contains a preference/decision/setting, store it as a short durable memory.
Default behavior (no tags)
- Be conservative about cross-topic mixing.
- If the user complains about topic bleed, suggest using the tags above.
Command-style cheat sheet responses
If the user sends /ctx or /context_def, respond with a short copy/pasteable cheat sheet:
- tags + one-line meaning
- two examples
Cross-platform / chat-surface notes
- Telegram slash commands cannot contain dashes.
- Use
/context_def(underscore), not/context-def.
- Use
- The tags themselves are just text; they work the same on Telegram/Discord/Slack/WhatsApp.
- If a surface auto-formats brackets, it’s fine—just keep the tags at the very beginning.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-phenomenoner-context-scope-tags": {
"enabled": true,
"auto_update": true
}
}
}Related Skills
openclaw-agent-optimize
Use when: you want to optimize an OpenClaw setup (cost/quality tradeoffs, model routing, context discipline, delegation, reliability) and you’re okay with a structured audit → options → recommended plan. Don’t use when: you want immediate config mutations without review, or the question is unrelated to OpenClaw operations. Output: a prioritized plan + exact change proposals (with rollback) if approved.
openclaw-agent-token-optimizer
DEPRECATED — duplicate listing. Please use the canonical "openclaw-agent-optimize" skill instead.
cron-worker-guardrails
Use when hardening OpenClaw cron workers (especially isolated agentTurn jobs) against quoting failures, brittle shell patterns, SIGPIPE false failures, and cwd/env drift. Output: a scripts-first hardening checklist + portable patterns.
context-clean-up
Use when: you suspect OpenClaw prompt context is bloating (slow replies, high cost, repeated transcript noise) and you want a ranked offender list + a reversible clean-up plan. Don’t use when: you want the assistant to apply fixes automatically, or you’re asking for unrelated troubleshooting. Output: an audit summary + 3–8 concrete fix steps + rollback notes (no automatic changes).
openclaw-agent-optimize-skill
DEPRECATED — duplicate listing. Please use the canonical "openclaw-agent-optimize" skill instead.