aidoc-flow

Save current plan and tasks to implementation file

**Specialist Claude Code subagents for the AI Doc Flow Framework's Specification-Driven Development (SDD) flow.**

Use this agent to review code changes and pull requests for quality, standards compliance, spec/test conformance, and acceptance-criteria verification. A READ-ONLY quality gate: it reports findings and a verdict, it never edits code itself. A downstream-heavy role built out in depth, since the success-story data shows review/test compression drives the largest gains.

Use this agent for CI/CD, build/test pipelines, deployment governance, and release readiness across the execution lane. Owns the path from merged code to deployed, observed, and verified release, including staging/prod gates and the post-deploy evidence loop.

Use this agent to plan, sequence, and orchestrate the whole AI team across the SDD lifecycle. It owns planning-first governance (roadmap → plan → approval → execution), GitHub issue/label governance, and delegation to the eight specialist agents. It drives the document lifecycle through the plugin's native SDD skills and is the PM seat the team plugs into.

Use this agent when decomposing, analyzing, and validating requirements across the SDD workflow. Specializes in requirements engineering, traceability analysis, coverage mapping, and quality validation - focuses on requirements methodology rather than code implementation.

Generate a phased implementation roadmap from a set of Architecture Decision Records, with dependencies, timeline, risk, and testing strategy. Use when coordinating delivery of multiple ADRs.

Create and manage Mermaid architecture diagrams as separate, traceable files linked to their parent SDD documents. Use when adding or migrating diagrams for BRD, PRD, ADR, SPEC, or IPLAN documents.

Scan a project's SDD artifacts and build a context model (inventory, traceability graph, workflow position, upstream candidates) to inform new document creation. Use before authoring an artifact in an existing project.

Audit an ADR - run declarative structural checks plus content review and produce a combined report for doc-adr-fixer. Use for ADR quality gating before SPEC.

Generate ADRs end-to-end from an upstream artifact (BRD/PRD/EARS/BDD), a prompt, or an IPLAN - detect input, scope the decision, generate, validate, and run the audit/fix cycle. Use to create or batch-create ADRs.