project-awareness
Build and maintain a three-layer project awareness system that gives Claude Code both the 30,000-foot architectural view AND the ability to drill into specifics — without exceeding context limits. Use this skill whenever: starting work on any project with 10+ source files; the user says "bootstrap", "set up awareness", "analyze this project", "build the map"; the user says "you're missing the big picture", "you don't understand the architecture", "you keep forgetting", "update the docs"; a CLAUDE.md exists with an [AWARENESS] section (this means the system is already bootstrapped — load and follow it); or when starting an Agent Teams session on any project (each agent must load its subsystem doc). This skill supersedes and works alongside deep-focus. Deep-focus is HOW you work carefully; project-awareness is HOW the project's knowledge is organized and maintained.
Target: Office & team collaboration. Automatically collected · unverified.
Excerpt from the author’s documentation
Extracted from a fixed version of SKILL.md in its original language. This is the author’s documentation; installation, compatibility and runtime results have not been verified by this site.
Read the original English text
Topics in the source
- Project Awareness: Three-Layer Context Architecture
- The Three Layers
- Layer 1: Constitution (CLAUDE.md — always loaded, 150-200 lines)
- Layer 2: Subsystem Docs (.claude/awareness/subsystems/.md — loaded on demand)
- Layer 3: Structural Map (.claude/awareness/structural-map.md — loaded on demand)
- Bootstrap Protocol (First Time Setup)
- Phase 1: Reconnaissance
- Phase 2: Subsystem Identification
You cannot hold an entire non-trivial codebase in one context window. Stop trying. Instead, maintain a structured knowledge system that gives you the 30,000-foot view always, subsystem details on demand, and a compressed structural map for navigation. The project's CLAUDE.md is the "hot memory." It loads on every session start. It contains: What the project IS (2-3 sentences) Module/subsystem inventory (name + one-line purpose each) Key architectural invariants (the rules that must NEVER be violated) Design tensions (known conflicts in the architecture, e.g., "block cache is 4K but AG headers are 512-byte sectors — every write to block 0 must use sector granularity") Routing table: which subsystem doc to load for which task Build/test commands The [AWARENESS] marker with metadata (last update, coverage stats) This document is the SINGLE MOST IMPORTANT artifact. If an agent does nothing else, reading this document should prevent the most common mistakes. Each major subsystem gets its own document. These are the "domain expert" docs. They contain: What the subsystem does and its public API Internal architecture (key data structures, state machines, flows) Invariants specific to this
This is an excerpt. Refer to the complete source and its current requirements.
- Source repository
- sshoecraft/ccenv
- Source file
- ccproject/SKILL.md
- Fixed revision
- e17f4107f24057f83d8727e1370f201ac6d7c255
Sources and evidence
This record was collected automatically. Its documentation, suitability and runtime behavior have not been manually verified.
Author: sshoecraft/ccenv · original repository metadata. Documentation review: Not reviewed. Last source check: 2026-11-10T21:01:31.165Z.
Categories are inferred from names, descriptions and paths. They do not establish compatibility.
License and attribution
MIT. Third-party content and copyright remain with their authors. Refer to the original license and terms.
Original license ↗Synchronization, copying and user-reported success do not mark a resource as runtime-tested.
Limitations and notes
- Collecting metadata does not mean this site has installed, adapted, runtime-tested or security-reviewed the resource.