Forge Intelligence

Handbook

Maturity roadmap

See Maturity state for the current snapshot and Maturity hierarchy for rollups. Per-intelligence uplift: Maturity heatmap. Gap remediation backlog: LMETA-LCDL gap report (run python3 tools/audit_lmeta_lcdl_gaps.py before…

Updated

Completed waves (A–J)

Wave Focus Outcome
A–E General spine, reasoning reference, life clusters Partial I1 promotion
F Life equalization Uniform I1_runtime floor — 297 allowlist, 39/39 eval pass
G Validator infra + honest reskilling I2 + spine A1 Contract validator dispatch; 10 I2 caps wired; 40 caps at A1 (reasoning + spine)
G2 Eval-normalized I1 One representative cap per intelligence with validate step; validator_exercised in eval status
H2 Agency integration case agency_weekly_rhythm — household, nutrition, personal-org under shared constraints
H1 Wellbeing integration case wellbeing_stress_recovery — A0 autonomy; escalation paths documented
H3a Society/risk case society_risk_navigation — civic, digital, consumer, emergency-readiness
H3b Home/mobility case home_mobility — repair triage + transit + domestic coordination
I Selective life A1 Low-risk life intelligences (household, nutrition, housing, mobility, personal-org)
J1 Platform routing capability_demand_router.yaml, implementation_bindings.yaml, capability_demand.py
J2 Benchmark corpus benchmarks/human_life/corpus_index.yaml + run_benchmark_corpus.py
J3 IRS/MO extensions +18 capabilities; index 315 total

Wave G2 deliverables

  • tools/wire_eval_validators.py + registries/eval_representative_caps.yaml
  • run_evaluation_suite.py — validator_exercised on representative cap
  • tests/test_eval_validator_wired.py

Wave H deliverables

  • registries/integration_cases.yaml — case_id → I2 caps (audit source of truth)
  • tools/wire_case_validators.py, tools/promote_case_i2.py
  • Case YAMLs under bundles/.../examples/cross_intelligence_*_case.yaml
  • Full-case tests in tests/test_integration_cases_full.py
  • audit_capability_maturity.py — i2_outside_known_cases gate

Wave I deliverables

  • tools/promote_wave_i_life_a1.py — A1_supervised on bounded low-risk life caps
  • Runtime docs: life A1 tier in INTELLIGENCE-NODE-RUNTIME.md

Wave J deliverables

Item Path
Demand router registries/capability_demand_router.yaml
Implementation bindings registries/implementation_bindings.yaml
Routing API orchestration/capability_demand.py
Benchmark index benchmarks/human_life/corpus_index.yaml
Benchmark runner tools/run_benchmark_corpus.py
Extension merge tools/merge_extension_packs.py

Realistic ceiling (v4.1 + extensions)

Metric Current
Indexed capabilities 315 (297 audited + 18 extension)
Allowlisted runnable 297 (audited v4.1)
I2 integrated (honest) 44 (union of integration cases)
A1 supervised spine + selective life (see MATURITY-STATE)
Eval suites 39/39 pass with validator exercised

Rule: Never promote I2 or A1 by registry sweep alone; always tie claims to executable evidence (validators, cases, audit gates).

Maturity vs autonomy

Technical maturity and autonomy are orthogonal (VOCABULARY.md). Wellbeing and high-risk life intelligences remain A0 even when case nodes reach I2.