Portfolio Development Progress

●Draft

Canonical baseline

The deterministic development-progress methodology is adopted as canonical under axodus-portfolio-development-progress/v1 following formal CTO acceptance. The historical 56.8% figure is retired as the current measurement and preserved solely for migration history. This score is a development evidence metric, not a production-readiness, financial-authority, or runtime execution claim.

Current portfolio development score

• Portfolio: 71.6% (canonical internal: 71.5633%)
• ACS: 61.8% (canonical internal: 61.8000%)
• Trading: 70.9% (canonical internal: 70.9000%)
• Repository evidence inventory: 18 repositories.
• Canonical scoring population: 15 nuclei at equal exact weight 1/15.
• Institutional remains included in the population.

The complete generated matrix, ACS and Trading traces, migration comparison, validation record, and hashes are maintained in the local Axodus coordination workspace under .instructions/portfolio_progress_*. The Documentation Overview is a downstream reader and never a scoring input.

The 6 November 2026 review generated a new repository-fingerprint manifest while preserving every capability state. Governance recorded local NEURONS implementation, artifact-verification, and same-host reproducibility evidence pending CTO risk and release decisions; the displayed portfolio, ACS, and Trading values therefore remain unchanged.

What the score measures

The score aggregates capability states across Research, Requirements and Scope, Architecture and Contracts, Implementation, Validation, Integration, Operational Readiness, and Production Readiness. Each dimension has a frozen v1 weight. Capability and program weights are explicit. Calculations use exact rational arithmetic internally; rounding occurs only when displaying the result.

Missing evidence is distinct from NOT_STARTED. Ambiguous or conflicting required evidence fails the calculation instead of being assigned a guessed percentage. Blockers affect only their declared capabilities and dimensions. Research, POCs, and provider evaluations do not automatically increase implementation or integration progress.

Evidence and limitations

The manifest records repository revisions, worktree fingerprints, capability states, evidence paths and hashes, blockers, methodology hash, and input-manifest hash. Reproduction does not require an LLM or network access.

All 18 repositories were inspected. AgentsAI, .github, and .github-private provide coordination or supporting evidence and have no standalone program weight. Institutional is scored from its repository evidence and remains in the 15-program denominator.

The current root coordination guide validator reports 65 marker/link defects, down from 69 after Governance submodule cleanup. They are classified as non-scoring coordination-quality defects under v1 because they are not frozen product capability acceptance criteria. Generated Documentation reports are outputs, not scoring inputs.

No production deployment, live trading, financial execution, wallet signing, treasury movement, settlement, payout, minting, provider adoption, or on-chain write is authorized by this score or its calculation.