margot-sandcastle-bridge
CommunityTurn research findings into approved tickets.
Software Engineering#linear#intent classification#audit logging#autonomous execution#github workflow#margot#hitl gate
AuthorCleanExpo
Version1.0.0
Installs0
System Documentation
What problem does it solve?
This Skill converts Margot deep research outputs into either actionable engineering work (via a HITL-approved Linear ticket) or safe informational logging, preventing lost research and avoiding accidental autonomous changes.
Core Features & Use Cases
- Intent classification with a safety confidence floor: routes only sufficiently confident, action-shaped research into action paths and logs the rest as informational.
- HITL gate before Linear issue creation: drafts the Linear ticket body and requires founder approval (Telegram 👍) prior to issue creation.
- Autonomy execution via existing poller + isolated runner: hands off accepted issues to the existing autonomy pipeline, which triggers sandcastle-runner using a high-isolation container path.
- Auditability and reversibility: records classification, drafting, acceptance, rejection, and expiry outcomes to append-only ledger files for later lessons/audits.
Quick Start
Use the margot-sandcastle-bridge workflow to classify a Margot research completion by intent, draft a founder-reviewed Linear ticket when appropriate, and route approved actions to isolated execution.
Dependency Matrix
Required Modules
None requiredComponents
Standard package💻 Claude Code Installation
Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.
Please help me install this Skill: Name: margot-sandcastle-bridge Download link: https://github.com/CleanExpo/Pi-Dev-Ops/archive/main.zip#margot-sandcastle-bridge Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
Agent Skills Search Helper
Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.