project:
name: embeddenator-agentic-dev
description: Multi-agent orchestration layer for developing and extending the embeddenator holographic VSA substrate
root_dir: ../.. language: Rust (core) + Python (orchestration)
tech_stack: sparse ternary vectors, VSA (bundle/bind/cleanup), engram encoding, Docker, CLI tools
license: MIT
agents:
- role: Planner
goal: Decompose features (ingest/extract/query/engram ops/scalability) into tasks
instructions_file: .github/instructions/planner_vsa.md
- role: RustEngineer
goal: Write/modify Rust code in src/ for SparseVec, EmbrFS, VSA ops
instructions_file: .github/instructions/rust_vsa_engineer.md
- role: PythonOrchestrator
goal: Improve orchestrator.py, build scripts, test_runner.py
instructions_file: .github/instructions/python_orchestrator.md
- role: Tester
goal: Write/maintain unit/integration/E2E tests (33+ coverage)
instructions_file: .github/instructions/tester_embeddenator.md
- role: VSAExpert
goal: Optimize ternary sparse vectors, bundle/bind ops, cosine cleanup, hierarchical chunking
instructions_file: .github/instructions/vsa_specialist.md
- role: DockerDevOps
goal: Maintain Dockerfile.tool, Dockerfile.holographic, OS engram builds
instructions_file: .github/instructions/docker_holographic.md
- role: Reviewer
goal: Enforce zero clippy warnings, production-grade quality, bit-perfect fidelity
instructions_file: .github/instructions/reviewer_embeddenator.md
task_tracker: tasking/tracker_vsa.yaml
workflows:
- feature_branch_prefix: feat/vsa-
- pr_require: Tester + Reviewer approval
- ci_integration: Reuse existing .github/workflows/ci-*.yml
custom_prompts:
- vsa_operation: "When implementing VSA ops, ensure sparsity <5%, ternary {-1,0,1}, algebraic correctness."
- engram_fidelity: "All extract operations must guarantee 100% bit-perfect reconstruction."