agentleFS
PhoenixNestPhoenixNest/AGENT-GLOBAL-BASE/core-component-00/CLAUDE.md

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

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
  1. Very specific:

    All paths relative to core-component-00/:

    line 49
  2. Distinctive:

    Run tests per-module from the module folder to avoid duplicate-package import collisions:

    line 66
  3. Very specific:

    Do NOT run all modules together with a single root-level pytest .

    line 76
  4. Very specific:

    All LLM-powered systems built in this workspace are bound by the ASGF framework — ratified via ADR-ASGF-001.

    line 101

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
CLAUDE.md

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

FieldDetail
NameDr. Elias Vance
Internal Codenamecore-component-00
RoleLaboratory Director — Core Component 00
Full Profilecrew/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

LayerModule FolderTypeHas Tests?
1 — What to writeframework/01-prompt-engineering/Knowledge baseNo
2 — How to structure itframework/02-context-engineering/Knowledge + FrameworkYes
3 — How to execute safelyframework/03-harness-engineering/Production FrameworkYes
4 — Where to get contentframework/04-retrieval-augmented-generation/Production FrameworkYes
5 — How agents cooperateframework/05-multi-agent-engineering/Production FrameworkYes

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/:

FileModulePurpose
framework/02-context-engineering/implementations/context_assembler.pyCEFour-slot context window assembly
framework/02-context-engineering/implementations/memory_store.pyCEEpisodic, semantic, procedural, working memory
framework/02-context-engineering/implementations/context_compressor.pyCELong-session compression
framework/03-harness-engineering/implementations/error_boundary.pyHETimeout, rate-limit, validation recovery
framework/03-harness-engineering/implementations/context_monitor.pyHEToken budget enforcement
framework/03-harness-engineering/implementations/tool_registry.pyHETool whitelists, call limits, dangerous task detection
framework/05-multi-agent-engineering/implementations/swarm_orchestrator.pyMAESwarm topology orchestration
framework/05-multi-agent-engineering/implementations/handoff_packet.pyMAEContext 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/ -v

Do 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.txt is aspirational, for architecture its fundamentals//patterns/ document but implementations/ 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 own pyproject.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 on sys.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/, numbered 00–05 (00-agent-systems-governance-framework/ through 05-multi-agent-engineering/). MCP servers, maintenance records, benchmarks, and remediation live under platform/ (unnumbered).


Where to Look

I need…Go to
Full lab overview + researcher profileREADME.md
Governing framework + complianceframework/00-agent-systems-governance-framework/
Synthesis of all five layersframework/00-agent-systems-governance-framework/CONCEPTS.md
Prompt patternsframework/01-prompt-engineering/
Context window architectureframework/02-context-engineering/
Safe model executionframework/03-harness-engineering/
RAG pipelinesframework/04-retrieval-augmented-generation/
Multi-agent / swarm systemsframework/05-multi-agent-engineering/
MCP server implementations (deployment surface)platform/model-context-protocol-servers/
Lab Director persona + crew rostercrew/
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 researchworkspace-root telescope/

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.