Pattern Proposal: Steering Block Workspace Bootstrapping

Summary

Injects standardized governance and architectural steering prompt blocks into target workspaces during environment bootstrap to ground external developer agents.

Candidate Pattern

• ID: steering-block-workspace-bootstrapping
• Classification: intelligence
• Maturity: established
• Type: implemented

Observed Signals

No specific signals listed.

Evidence

Supporting Evidence

• src/cas_runner/bootstrap.py (merge_cas_governance_block, bootstrap, init_workspace) - Logic inserting and updating CAS steering blocks into workspace agent files
• AGENTS.md - Workspace agent guidelines file managed by CAS steering

Scope Assessment

• Evidence Strength: high
• Confidence: high
• Applies to: ideas-to-life-cas

Rationale

• Intent: Ensure that localized AI tools (such as Claude Code, GitHub Copilot, Gemini CLI) adhere to system architecture guidelines from project initialization.
• Context: External coding assistants in developer workspaces lack awareness of project-specific architecture rules unless explicitly prompted.
• Force: Developers forget to feed architectural guidelines into their localized AI coding prompts.
• Force: Manual copying of prompt guidelines leads to divergence across repositories.

Traceability

Artifact source: proposed-patterns