Skip to content

CORE CLI Commands Table

This document provides a clean, accurate, up-to-date command table for the core-admin CLI. It reflects:

  • all commands that actually exist in your current CLI tree,
  • grouped logically,
  • with short, actionable descriptions,
  • PLUS a separate section for Planned / Experimental commands that appear in design docs but do not exist in this version.

This avoids ambiguity and keeps developers aligned with reality.

✅ 1. Real Commands (Implemented in src/body/cli/commands/)

These commands are shipped and available today.

1.1. check — Validation & Governance

Command Description
check audit Run the full Constitutional Audit (Mind checks).
check status Show project health summary.
check validate Run validation pipeline (formatting, linting, syntax, tests) (if present in your version).

1.2. develop — Autonomous Crate Generation

Command Description
develop feature "…" Generate a crate for a new feature using governed AI.

1.3. fix — Automated Self-Healing

Command Description
fix ids --write Fix or assign # ID: tags.
fix code-style --write Run Black + Ruff formatting/linting.
fix docstrings --write Generate or correct docstrings.
fix headers --write Ensure file headers match standards.
fix purge-legacy-tags --write Remove outdated capability tags.
fix all --write Apply all safe remediations.
fix all --dry-run Show what would be changed.

(Commands are backed by services under src/features/self_healing/.)

1.4. inspect — Introspection & Diagnostics

Command Description
inspect command-tree Display the full CLI command tree.
inspect project Show structural info (domains, capabilities, drift, etc.) (if present in your version).

1.5. manage — Knowledge, Database, Proposals

Command Description
manage database sync-knowledge Rebuild Knowledge Graph, symbols, capabilities.
manage proposals new "…" Create a new constitutional proposal.

1.6. mind — Governance Utilities

Command Description
mind show-policies List loaded governance policies (if implemented).
mind show-domains Inspect domain boundaries (if implemented).

1.7. run — Execution Utilities

Command Description
run process-crate <id> Process a specific crate (only if present).

1.8. search — Search Symbols & Capabilities

Command Description
search symbol "…" Search for symbol names via Knowledge Graph.
search capability "…" Search for capabilities.

1.9. secrets — Local Secrets Management

Command Description
secrets init Initialize developer secrets store.
secrets set NAME Set a secret key locally.

📦 2. Additional Commands (Exist in Tree but Vary by Version)

These commands appear in your directory tree, but their exact availability depends on how your CLI wiring is configured.

Command Description
enrich symbols Enrich symbol metadata (if wired).
enrich capabilities Recompute capability metadata (if wired).
submit proposal <path> Submit a governance proposal (depending on your version).
submit crate <id> Submit crate metadata (rarely used in 2025 design).

These are real modules, but may not be registered in your installed CLI.

⚠️ 3. Planned / Experimental (Not Available in This Version)

These commands appear in old docs, design notes, or roadmap discussions, but are not implemented today. They are included here to prevent confusion.

Command (Not Implemented) Why It’s Listed
coverage check Planned advanced coverage tooling (self-healing integration).
coverage remediate Planned coverage-aware remediation pipeline.
manage proposals approve Advanced governance workflow (A4+).
submit changes Deprecated early idea (no longer part of A1–A3 roadmap).
daemon start Historical design reference — no daemon exists in current CORE.
daemon stop Same — deprecated concept, removed.
develop refactor Planned future A3 governed refactoring command.
develop test Planned test-first generation (superseded by A2 agent planner).

These entries help orient contributors without implying they are available.

🧭 4. Command Map (Mind → Body → Will)

graph TD;
    subgraph Mind[Governance]
        A[check audit]
    end
    subgraph Body[Execution]
        B[fix ...]
        C[manage database]
        D[inspect ...]
        E[search ...]
    end
    subgraph Will[Agents]
        F[develop feature]
    end

    F --> B
    F --> A
    B --> A
    C --> A
    D --> A

This diagram shows:

  • Will (agents) trigger Body (validation) and Mind (audit).
  • Body commands often run Mind checks.
  • Mind never calls Will directly.

🧩 5. Summary

  • Section 1: Commands that definitely exist in the current CORE installation.
  • Section 2: Commands that exist in the repo but may not be CLI-registered.
  • Section 3: Planned/experimental commands — not implemented but appear in design materials.

This structure ensures developers have:

  • clarity,
  • correctness,
  • traceability,
  • and no ambiguity between real, maybe, and not implemented.

If the CLI changes, update this document immediately — the CLI is the authoritative gateway to CORE’s governed autonomy.