AGENT-GLOBAL-BASE / core-component-00
CLAUDE.md2 starsChanged 2 months agoApache-2.0View on GitHub
Entry point for the Core Component 00 applied research laboratory. Read this before any CC-00 work. CC-00 is the organisation's centralised LLM engineering laboratory and the foundational dependency for every agent-powered system built here. It houses five engineering modules plus a governing meta-layer (ASGF), production-grade Python implementations, and active research programmes. core-component-00/ holds the stack's reference implementations and the MCP servers. As of FY2026 Q3, the lab is staffed beyond the Director: 4 Research Engineer FTEs cover the four production-grade…
- Installs packages
In this repository27 other files it gives its agents.
- AGENTS.mdAGENTS.md
- academic-neural-unit-00/CLAUDE.mdCLAUDE.md
- CLAUDE.mdCLAUDE.md
- company/CLAUDE.mdCLAUDE.md
- studio/CLAUDE.mdCLAUDE.md
- telescope/CLAUDE.mdCLAUDE.md
- activate-org-agentSkill
- android-engineeringSkill
- audio-engineeringSkill
- backend-engineeringSkill
- cross-platform-engineeringSkill
- cyberspace-securitySkill
- data-analyticsSkill
- engineeringSkill
- frontend-engineeringSkill
- game-developmentSkill
- ios-engineeringSkill
- legalSkill
- live-operationsSkill
- llm-engineeringSkill
- localizationSkill
- product-designSkill
- product-managementSkill
- quality-assuranceSkill
- recruitmentSkill
- technology-strategySkill
- visual-arts-and-animationSkill
The file
We split this CLAUDE.md into the rules it gives an agent
4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 49
All paths relative to core-component-00/:
- Distinctive: line 66
Run tests per-module from the module folder to avoid duplicate-package import collisions:
- Very specific: line 76
Do NOT run all modules together with a single root-level pytest .
- Very specific: line 101
All LLM-powered systems built in this workspace are bound by the ASGF framework — ratified via ADR-ASGF-001.
≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
core-component-00/ — CC-00 LLM Engineering Laboratory
Entry point for the Core Component 00 applied research laboratory. Read this before any CC-00 work.
What This Is
CC-00 is the organisation's centralised LLM engineering laboratory and the foundational dependency for every agent-powered system built here. It houses five engineering modules plus a governing meta-layer (ASGF), production-grade Python implementations, and active research programmes.
core-component-00/ holds the stack's reference implementations and the MCP servers.
Laboratory Director & Crew
| Field | Detail |
| Name | Dr. Elias Vance |
| Internal Codename | core-component-00 |
| Role | Laboratory Director — Core Component 00 |
| Full Profile | crew/director/elias-vance/agent/profile.md |
As of FY2026 Q3, the lab is staffed beyond the Director: 4 Research Engineer FTEs cover the four
production-grade modules (Context, Harness, RAG, Multi-Agent). Full roster and activation
protocol: crew/README.md and crew/CLAUDE.md.
The Five-Module Engineering Stack
| Layer | Module Folder | Type | Has Tests? |
| 1 — What to write | framework/01-prompt-engineering/ | Knowledge base | No |
| 2 — How to structure it | framework/02-context-engineering/ | Knowledge + Framework | Yes |
| 3 — How to execute safely | framework/03-harness-engineering/ | Production Framework | Yes |
| 4 — Where to get content | framework/04-retrieval-augmented-generation/ | Production Framework | Yes |
| 5 — How agents cooperate | framework/05-multi-agent-engineering/ | Production Framework | Yes |
ASGF (Agent Systems Governance Framework) is the meta-layer above all five — not a sixth module. See
framework/00-agent-systems-governance-framework/.
Key Production Implementations
All paths relative to core-component-00/:
| File | Module | Purpose |
framework/02-context-engineering/implementations/context_assembler.py | CE | Four-slot context window assembly |
framework/02-context-engineering/implementations/memory_store.py | CE | Episodic, semantic, procedural, working memory |
framework/02-context-engineering/implementations/context_compressor.py | CE | Long-session compression |
framework/03-harness-engineering/implementations/error_boundary.py | HE | Timeout, rate-limit, validation recovery |
framework/03-harness-engineering/implementations/context_monitor.py | HE | Token budget enforcement |
framework/03-harness-engineering/implementations/tool_registry.py | HE | Tool whitelists, call limits, dangerous task detection |
framework/05-multi-agent-engineering/implementations/swarm_orchestrator.py | MAE | Swarm topology orchestration |
framework/05-multi-agent-engineering/implementations/handoff_packet.py | MAE | Context Handoff Protocol |
Running Tests (PowerShell)
Run tests per-module from the module folder to avoid duplicate-package import collisions:
# From core-component-00/
pytest framework/02-context-engineering/testing/ -v
pytest framework/03-harness-engineering/testing/ -v
pytest framework/04-retrieval-augmented-generation/testing/ -v
pytest framework/05-multi-agent-engineering/testing/ -vDo NOT run all modules together with a single root-level pytest . — this causes import conflicts.
Parallel workers: pytest -n <N> — keep N ≤ 10 on this machine (i9-13900H, 14 cores).
Environment Notes (Windows)
- RAG's
requirements.txtis aspirational, for architecture itsfundamentals//patterns/document butimplementations/does not yet build — install from it only when actually implementing one of those pieces, not to run RAG's existing test suite (that only needs the module's ownpyproject.toml,pytest-only):pip install -r framework/04-retrieval-augmented-generation/requirements.txt python -m spacy download en_core_web_sm - GPU: RTX 4060 (8 GB GDDR6) supports CUDA — always verify
torch.cuda.is_available()before assuming GPU is available. - Import path: Tests use
from implementations.<module>after inserting the module root onsys.path. Run from the module folder, not the workspace root.
ASGF Governance (Mandatory)
All LLM-powered systems built in this workspace are bound by the ASGF framework — ratified via
ADR-ASGF-001. Build new systems on CC-00 patterns; do not invent ad-hoc approaches.
Governing documents: framework/00-agent-systems-governance-framework/governance/
Layout note: All five engineering modules plus ASGF now live under
framework/, numbered00–05(00-agent-systems-governance-framework/through05-multi-agent-engineering/). MCP servers, maintenance records, benchmarks, and remediation live underplatform/(unnumbered).
Where to Look
| I need… | Go to |
| Full lab overview + researcher profile | README.md |
| Governing framework + compliance | framework/00-agent-systems-governance-framework/ |
| Synthesis of all five layers | framework/00-agent-systems-governance-framework/CONCEPTS.md |
| Prompt patterns | framework/01-prompt-engineering/ |
| Context window architecture | framework/02-context-engineering/ |
| Safe model execution | framework/03-harness-engineering/ |
| RAG pipelines | framework/04-retrieval-augmented-generation/ |
| Multi-agent / swarm systems | framework/05-multi-agent-engineering/ |
| MCP server implementations (deployment surface) | platform/model-context-protocol-servers/ |
| Lab Director persona + crew roster | crew/ |
| Research reports (CC-00 engineering + LLM research) | telescope/ |
| Maintenance operations log (servers, dependencies, MCP infra) | platform/maintenance-records/ |
| Enterprise benchmark assessments (module vs. industry practice) | platform/benchmarks/ |
| Remediation execution log (P0/P1 benchmark findings) | platform/remediation/ |
| Cross-cutting / workspace-wide research | workspace-root telescope/ |
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

