CtrlK
BlogDocsLog inGet started
Tessl Logo
PluginPublic

jbaruch/nanoclaw-trusted

Rules for trusted NanoClaw groups. Shared memory, session bootstrap, cross-group memory updates. Loaded for trusted and main containers only.

Quality

92%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

Overview
Quality
Evals
Security
Files

context-bootstrap-bg-agents.mdrules/

alwaysApply:
No
applyTo:
** — when launching a background Agent, including prompts supplied by a skill

Context Bootstrap for Background Agents

When launching a background Agent, prepend the prompt with these workspace-context lines.

Required preamble lines

  • Workspace: /workspace/group/ (your files), /workspace/ipc/ (messaging).
  • Send results via mcp__nanoclaw__send_message.
  • Telegram HTML: <b>bold</b>, <i>italic</i>, • bullets. No markdown.

When to omit

If the parent prompt already includes equivalent context (e.g., a Skill that injects it), don't duplicate the preamble. Duplication confuses the sub-agent about which copy is authoritative.

Tools available to sub-agents

  • mcp__nanoclaw__send_message — outbound IPC to chat
  • mcp__nanoclaw__react_to_message — reaction on the originating message
  • The full Skill() tool surface — sub-agents can invoke any installed skill

README.md

tile.json