Skip to main content

Core Concepts

Command notation: Commands in this guide are shown in Claude Code format (/wire:*). If you are using Gemini CLI, drop the /wire: prefix and replace colons with spaces — e.g., /wire:requirements-generate my_project becomes /dp requirements generate my_project.

Self-contained command architecture​

Every /wire:* command is a single, self-contained file — the command file is the complete workflow specification. There is no separation between a discovery layer and a logic layer. In Claude Code, these are .md files distributed as a plugin; in Gemini CLI, .toml files distributed as an extension.

Each command file contains the full workflow inline — from 100 lines for a simple review command to over 1,500 lines for dbt generation.

The artifact lifecycle​

Every artifact produced by the framework follows three gates:

  • Generate: AI produces the artifact from upstream inputs and templates
  • Validate: Automated checks run (naming, test coverage, completeness, etc.)
  • Review: You or the client approves the artifact

An artifact should not progress until all three gates are passed. Downstream artifacts check upstream readiness before they generate.

The precondition gate​

Since v4.0.0. Every -generate/-validate/-review command auto-delegates to a shared utility, precondition_gate.md, before doing anything else. It reads the command's declared preconditions from its own front-matter — a static list (e.g. "data_model.review must be approved"), or the dynamic sentinel for the handful of artifacts (mockups, pipeline_design, data_model, data_quality, dashboards, deployment, training, documentation) whose correct precondition genuinely differs by release type. A dynamic precondition resolves at runtime from the current release's wire/release-types/<type>.yaml — the same file Autopilot reads to resolve execution order.

If the precondition isn't met, the command blocks by default. You can override it, but only explicitly — the gate asks for your name and a reason, and records both in status.md's precondition_overrides and in execution_log.md as an override result. This makes "I skipped a step on purpose" a visible, attributable decision rather than something that just silently happened.

Git branching​

/wire:new enforces a mandatory branch check. If you run it while on main or master, the framework will stop and ask you to create a feature branch before any project files are created. It suggests feature/{folder_name} but you can choose your own name.

This ensures all release work lives on a branch that can be reviewed via pull request before merging.

The status file​

Each release has a status.md file at .wire/releases/<release-folder>/status.md. This is the running instance of the delivery process — created by /wire:new when you select a release type, and updated by every subsequent command. It has two roles:

  1. Human-readable: release overview, notes, blockers, and session history
  2. Machine-readable YAML frontmatter: the instantiated process definition — which artifacts are in scope, which gates have been passed, and what comes next

The framework updates status.md automatically after each command.

The execution log​

Each project maintains an execution_log.md file that records a timestamped entry for every command that changes state:

| Timestamp | Command | Result | Detail |
|-----------|---------|--------|--------|
| 2026-02-22 14:40 | /wire:requirements-generate | complete | Generated requirements spec (3 files) |
| 2026-02-22 15:12 | /wire:requirements-validate | pass | 14 checks passed, 0 failed |
| 2026-02-22 16:00 | /wire:requirements-review | approved | Reviewed by Jane Smith |

The chain of derivation​

Each artifact constrains the next. By the time the AI generates LookML, the dimension names, measure definitions, and join paths are fully determined by upstream artifacts — there is no room for improvisation.

Specialist agents​

As of v3.9.4, Wire commands auto-delegate to one of fourteen specialist subagents — a dbt-developer agent that only knows dbt conventions, a qa-agent that is a pure critic with no generation responsibility, and so on. This happens transparently when you run individual commands. To batch-delegate all pending work across an entire release, use /wire:delegate <release-folder>.

See Wire Agents for the full agent roster and how delegation works.

Research persistence​

When the AI performs technical research during a session, it automatically saves structured summaries to .wire/research/sessions/YYYY-MM-DD-HHMM/summary.md. The engagement-context skill checks these saved summaries when loading context — if a relevant prior finding exists, it is surfaced rather than re-running the same research.

This means:

  • Cross-release knowledge carries over: research done during the discovery release is available when working on the delivery release
  • Re-starting a session doesn't lose context: prior technical findings are always available
  • Less AI context consumed: the AI reads a condensed summary instead of re-running the same web searches