Skip to content

TODO.md

Latest commit

 

History

History
283 lines (221 loc) · 11.3 KB

TODO.md

File metadata and controls

283 lines (221 loc) · 11.3 KB

GEO-INFER — TODO & Release Roadmap

Last Updated: 2026-02-25
Current Version: 0.2.0 (Beta)
Repository: ActiveInferenceInstitute/GEO-INFER

📋 Release Criteria (Universal Gate)

Every version release MUST satisfy ALL of the following before tagging:

Category Criterion Verification Required
Quality All tests pass uv run python GEO-INFER-TEST/run_unified_tests.py 0 failures
Quality No stub/placeholder code grep -rn "placeholder|NotImplementedError|stub" --include="*.py" GEO-INFER-*/src/ 0 results¹
Quality No illegitimate pass stubs Grep ^ pass$ excluding __init__, except, abstract 0 results
Quality Type hints complete mypy --strict on core modules 0 errors
Quality Formatting & lint black --check, isort --check, ruff check Clean
Docs Every module has README.md + AGENTS.md find GEO-INFER-*/README.md | wc -l 44 each ✅
Docs Every module has SKILL.md (Claude Code) python GEO-INFER-TEST/validate_skills.py 0 errors ✅
Docs No stale dates Grep for old dates excluding CHANGELOG/TODO 0 results ✅
Docs CHANGELOG.md updated Manual inspection Entry present ✅
Testing All 44 modules have ≥4 test files find GEO-INFER-*/tests -name "test_*.py" | wc -l ≥176 ✅ (416)
Testing Coverage ≥80% per module pytest --cov --cov-fail-under=80 All pass
Testing Property-based tests ≥10 modules Grep @given|hypothesis ≥10 ✅ (35)
Arch PEP 8 package names No unexpected package dir casing in src/ 0 ✅ (all 44 packages normalized to geo_infer_<module>)
Arch Graceful dependency degradation import geo_infer_act without optional deps No ImportError ✅
Arch H3 v4 API only No legacy h3.geo_to_h3 calls 0 ✅

¹ Excludes legitimate uses: SQL parameter placeholders, HTML placeholder= attributes, fallback geometries, docstring references.

🔖 Version Roadmap

v0.2.0 — Current Release ✅

Released: 2026-02-25 | Theme: Documentation parity, placeholder elimination, LOG integration

Completed items (click to expand)

Source Code Remediation (All resolved)

  • IOT: Real paho-mqtt MQTT handlers
  • AG: USDA soil health, IPCC Tier 1 carbon, FAO-56 water usage
  • ACT: Dirichlet/categorical generative model, real belief trajectory plotting
  • BAYES: Real ELBO, Jeffreys/reference/unit-info priors, Cholesky LKJ multilevel, TFP GP+MH
  • ART: Real mplcursors interactivity
  • SEC: Uses calculated_confidence (not hardcoded)
  • LOG: Verified clean (KMeans, Dijkstra, PuLP — 0 placeholders)
  • RISK: Verified clean (Cholesky, Moran's I, Monte Carlo — 0 placeholders)
  • COMMS: All placeholders resolved (PyJWT/HMAC, subscriber registry)
  • NORMS, ECON, EXAMPLES, AGENT, PEP: All addressed

Documentation & Infrastructure

  • All stale dates (2026-01-26, 2026-02-17) updated to 2026-02-25
  • CHANGELOG.md v0.2.0 entry added
  • pyproject.toml: License, URLs, version all correct
  • CI: .github/workflows/ci.yml exists
  • LOG __init__.py: 14 lazy exports + submodule access
  • TRANSPORT: BPR microsimulation, EWMA traffic forecast
  • SPM: Real time-series explorer (mean±SD + residuals)
  • Cross-module integrations: LOG↔TRANSPORT (emissions), LOG↔ECON (logistics)
  • 45 SKILL.md files deployed (1 root + 44 modules) with Examples, Guidelines, Integrations
  • validate_skills.py added to GEO-INFER-TEST for CI validation
  • SKILL.md signposted in 44 README.md nav bars + 44 AGENTS.md footers

v0.3.0 — "Test Green" Release

Target: March 2026
Theme: Fix all test failures, achieve ≥80% coverage in core modules

Test Fixing (23/47 categories currently passing → target: 47/47)

  • SPM: Fix coordinate validation — synthetic test data generates lat/lon outside ±90°/±180° bounds
  • SPM: Fix fit_glm import (NameError: name 'fit_glm' is not defined)
  • SPM: Fix SPMResult.cov_beta attribute error in ACT integration
  • ACT: Resolve jax/tensorflow_probability/oryx upstream dependency breakage
  • DATA: Investigate and fix test failures (38.6s runtime, details in test-results/DATA_results.xml)
  • ART: Fix test failures (29.3s runtime — likely visualization backend issues)
  • ANT: Fix test failures (213s runtime — algorithmic convergence timeout)
  • RISK: Fix test failures (9.9s — assertion/bounds issues)
  • OPS: Fix test failures (22s — monitoring integration)
  • LOG: Fix test_module_structure — hasattr(geo_infer_log, 'api') fails ✅ Fixed
  • NORMS, ECON, SPACE, SIM, MATH, BIO, HEALTH, SEC, AGENT, EDU: Investigate remaining failures

Coverage Targets

Module Group Current Target
Core Analytical (MATH, ACT, BAYES, SPM) Unknown ≥80%
Infrastructure (SPACE, TIME, DATA, API) Unknown ≥80%
Domain (all others) Unknown ≥60%

Remaining Source Code Items (0 open)

File Issue Priority Status
NORMS/.../compliance_tracking.py Simplified evaluation logic Medium ✅ Resolved
PLACE/.../unified_backend.py Fallback placeholder geometries Low ✅ Resolved
SPACE/.../visualization_engine.py 4 placeholder monitoring data points Low ✅ Resolved

v0.4.0 — "Full Coverage" Release

Target: May 2026
Theme: 90%+ coverage in core, complete documentation, strict typing

Code Quality

  • Mypy strict mode passing in all 44 modules
  • Test coverage ≥90% in Core Analytical modules (MATH, ACT, BAYES, AI, AGENT, COG, SPM)
  • Test coverage ≥80% in all Domain modules
  • Performance benchmarks for spatial operations (H3 indexing, geodesic calculations)

Documentation

  • Sphinx API documentation auto-generated for all 44 modules
  • Jupyter notebook tutorials for each domain category
  • "From Data to Active Inference" end-to-end tutorial
  • Docstring coverage ≥95% (enforced via pydocstyle)

Module Completions (Alpha → Beta)

  • SPM: Complete GLM spatial implementation
  • ANT: Verify ACO/PSO/ABC convergence
  • SIM: Real Mesa-based agent simulation
  • COG: Complete cognitive modeling (attention, memory, trust)

Infrastructure

  • Automated coverage reporting via Codecov
  • Pre-commit hooks (Black, isort, ruff, mypy)
  • Docker base image for development

v0.5.0 — "Domain Completions" Release

Target: August 2026
Theme: All domain modules reach Beta status

Module Current Issue Target
SEC Threat detection incomplete Anomaly-based intrusion detection
OPS Monitoring incomplete Prometheus/Grafana integration
METAGOV DAO mechanisms incomplete Real governance mechanisms
TRANSPORT Traffic models partial Full graph-based network flow
EMERGENCY Resource deployment incomplete Linear programming allocation
ENERGY LCOE benchmarking incomplete Complete techno-economic analysis
CIV STEW-MAP partial Full participatory mapping
REQ P3IF partial Complete traceability matrix

New Capabilities

  • API: Complete GraphQL schema for all 44 module endpoints
  • APP: Deploy reference dashboard application
  • PLACE: H3-based place-shedding and catchment area analysis

v1.0.0 — "Production Release"

Target: November 2026
Theme: All release criteria met, all 44 modules at Beta+

  • ALL universal release criteria gates pass
  • All 44 modules at Beta or higher
  • Zero placeholder/stub/fake/mock in source code
  • Full Sphinx docs at geo-infer.readthedocs.io
  • Semantic versioning in all 44 pyproject.toml files
  • Security audit for SEC and API modules
  • Performance benchmarks published (SPACE, TIME, BAYES)
  • CC BY-NC-SA 4.0 license compliance verified

🗂️ Module Status Registry (2026-02-25)

✅ Beta+ Modules (20/44) — No blockers

Module Status Notes
MATH Beta —
ACT Beta Placeholders resolved
BAYES Beta GP+MH, ELBO, priors all real
SPACE Beta (H3 v4) —
IOT Beta Real paho-mqtt
API Beta —
AG Beta USDA/IPCC/FAO-56
HEALTH Beta —
BIO Beta —
CLIMATE Beta —
FOREST Beta —
COMMS Beta All channels implemented
APP Beta —
ART Beta Real mplcursors
PLACE Beta (H3 v4) —
INTRA Beta —
GIT Beta —
TEST Stable —
EXAMPLES Beta Needs e2e tutorial
LOG Beta Verified clean, lazy exports

🟡 Alpha Modules (24/44) — Need work for v0.5.0

Module Blocker Target
AI None known v0.4.0
COG Incomplete attention/memory v0.4.0
AGENT None (telemetry resolved) v0.3.0
SPM GLM incomplete, test fixtures v0.4.0
TIME None known v0.3.0
DATA None known v0.3.0
SEC Threat detection incomplete v0.5.0
OPS Monitoring incomplete v0.5.0
METAGOV DAO incomplete v0.5.0
ECON None (auction resolved) v0.3.0
RISK None (verified clean) v0.3.0
ENERGY LCOE incomplete v0.5.0
WATER None known v0.3.0
TRANSPORT Traffic models partial v0.5.0
MARINE None known v0.3.0
EMERGENCY Deployment opt incomplete v0.5.0
EDU None (pass in template string) v0.3.0
SIM Mesa integration incomplete v0.4.0
ANT Convergence not verified v0.4.0
CIV STEW-MAP partial v0.5.0
PEP None (TODO → roadmap note) v0.3.0
ORG DAO incomplete v0.5.0
NORMS Simplified compliance eval v0.3.0
REQ P3IF partial v0.5.0

📊 Current Metrics

Metric Value Target Status
Test categories passing 23/47 47/47 🟡
Placeholder references (source) ~23 0 🟡 (most are legitimate)
Test files 421 ≥176 ✅
Unit test files 318 ≥44 ✅
Integration test files 50 ≥44 ✅
Hypothesis test refs 35 ≥10 modules ✅
README.md files 44 44 ✅
AGENTS.md files 44 44 ✅
Source files 860 — —
Source lines 297,360 — —
H3 legacy calls 0 0 ✅
Pass stubs 0 0 ✅
Stale dates in docs 0 0 ✅
Modules at Beta+ 20/44 44/44 🟡

Quick Verification

# Run all tests
uv run python GEO-INFER-TEST/run_unified_tests.py 2>&1 | tail -20

# Check remaining placeholder refs
grep -rn "placeholder\|stub\|fake" --include="*.py" GEO-INFER-*/src/ | \
  grep -v "abstractmethod\|__pycache__\|SQL\|HTML\|input\|fallback\|docstring" | wc -l

# Documentation completeness
find GEO-INFER-*/README.md | wc -l  # 44
find GEO-INFER-*/AGENTS.md | wc -l  # 44

# H3 legacy check
grep -rn "h3.geo_to_h3\|h3.h3_to_geo\|h3.k_ring" --include="*.py" . | wc -l  # 0

This TODO is a living document. Governed by the PAI Algorithm — all criteria are binary, testable, and verifiable.