dotscope 1.5.5

Agents see files. You see architecture. dotscope gives agents the architecture.

The Architectural Control Plane for AI Agents

Your agent writes code that compiles, passes tests, and breaks production. The agent sees files. You see architecture. dotscope closes that gap.

The Open-Source Architectural Control Plane

Dotscope is a high-performance headless telemetry engine built natively around the standard Model Context Protocol (MCP). It operates silently, mapping the topological physics of your codebase and transmitting structural constraints and contextual boundaries directly to your AI agent of choice (Claude, Cursor, Windsurf, Cline).

pip install dotscope
dotscope serve --headless

---

Strict Execution Taxonomy

Dotscope is engineered for uncompromising speed, operating strictly across decoupled bounds ensuring zero data drift between human CLI inputs and Agent MCP tool queries:

• <Engine>: Resolves abstract scope mathematics (e.g. auth+payments-tests&api@context) down to concrete file constraints with zero side effects.
• <Workflows>: Deploys multi-stage sequence migrations and automated ingest cycles that silently reverse-engineer architectural logic from raw Git history.
• <UX>: Formats headless semantic JSON outputs perfectly into FastMCP standard payloads and terminal outputs for instantaneous debugging scenarios.

$ dotscope ingest

  Analyzing dependency graph...
  Mining git history...
  Discovering conventions...

  Discoveries:
  - version.py and environment.prod.ts always change together
  - workflow-edit-dialog.component.ts and models.py are tightly coupled

  Validation (49 commits backtested):
  - Overall recall: 78%
  - Token reduction: 67% (1.3M → 437K avg)

  Output: 3 .scope files written.

One MCP tool call. The agent gets the relevant code, its dependency neighborhood, implicit contracts from git history, convention rules, swarm lock status, and action hints. One call, not five.

By executing dotscope ingest and mining your git history, dotscope statically learns from every commit. Files agents consistently need get ranked higher. Architectures are mapped historically. Conventions that hold get enforced harder. Rules that get overridden get quieter. Recall starts at 78% and climbs past 91%.

pip install dotscope && dotscope serve

Zero dependencies. Open Source. MIT.

How It Works · Scope Files · Agent Instructions · MIT